Data + AI Summit Keynote Thursday

[Music] please welcome oligoti hello everyone super super excited to be here we actually have now online 75 000 people worldwide that are actually listening in to this live stream [Music] lake house IQ is just like a dream to me just because of all the capabilities it has thank you amazing to learn about the implementation [Music] limitations database is going in and we're looking forward to it Unity catalog has been really important to us lake house Federation and monitoring yeah they make it even better foreign having fun with data it's not about educating yourself on what you're doing today it's about where the world is going [Music] please welcome to the stage director of machine learning Brooke winning [Music] don't let me good morning to everyone joining us here in person and good morning good afternoon and good evening to those of you joining us here virtually now I know it's early in the morning but please make some noise if you're enjoying day to AI Summit that was so good I'm not going to make you do that again data and AI Summit is truly the Highlight event of the year I love that it connects data teams from around the from around the globe and truly builds a community I first attended what was then known as spark Summit in 2016. and a few years later I was on this very stage demoing the latest features of Apache spark 3.0 I'm thrilled to be here with all of you today as a bit of housekeeping the Keynotes will run until 11AM there are no scheduled breaks but if you do need to get up we ask that you'd be respectful of our presenters and those around you and to do so in between sessions before we can get into the Keynotes I want to thank our partners without them this event truly would not be possible please put your hands together to thank our partners and I want to do a special shout out to Accenture avanade AWS Microsoft and prophecy please visit them in the expo hall after the Keynotes to kick off data in the eye Summit oops to kick off data and AI Summit we launched the so you think you can hack hackathon a focus on either building llm-powered applications or leveraging spark connect and we had both an in-person and a virtual track and over 800 participants participated in this hackathon so without further Ado I want to congratulate all of the finalists you see up on the slide here in particular I'm going to do a special shout out to the Grand Prize winners in each category schema spy who built it a system to navigate database schemas with AI assistance and in the virtual category data DM who built a data assistant to help you load transform visualize your data all without a single line of code and they actually open sourced it please congratulate all the finalists that you see here [Applause] now on to the Keynotes our mission is to democratize data and AI but that mission is not that of databricks alone it is a shared mission in the data and AI community in our keynote lineup today reflects that shared mission only a date and I Summit will you hear from the creators of the hottest open source projects Visionary academics and Silicon Valley Legends you will not want to leave the Keynotes early we will be closing the Keynotes with a fireside chat with Ali godzi and Eric Schmidt former CEO of Google I hope you all are as excited as I am to be here today [Applause] as Ali mentioned in his keynote yesterday technology Innovations don't become revolutions until they're truly democratized Mark Andreessen is no stranger to this through creating the Mosaic and Netscape web browsers Mark truly democratized the internet something that all of us rely on for our day-to-day lives 12 years ago he coined the term software is eating the world and a few weeks ago he published a bullish opinion on why he thinks AI will save the world without further Ado please join me in welcoming Mark and driesen and Ali godzi to the stage [Applause] [Music] [Applause] [Music] awesome thanks so much Mark thanks for doing this good morning good morning everybody can we get a hands up for those of you who've read his essay on why AI is saving the world foreign yeah there's quite a few people maybe we can start actually I'm curious so you started the first web browser mosaic um and we acquired Mosaic ml yesterday yes what do you think about that you know if the neighbors can get reused it should get reused for this so I'm excited awesome awesome so maybe maybe starting with that actually I'm kind of curious so you know you've been part of many many internet revolutions you've seen it you've had like the inside you know scoop first the browser the internet then social networking actually I don't know if everyone knows this but you were on the on the board of Facebook since 2007 2007. so you saw this whole thing go from like just the place where you check out Trends to you know this this thing that it is today and now ai is happening so why did you write this uh article yeah well so so AI is one of these things ai's been around right conceptually for a long time right one of the interesting things I think I find interesting about it is it was actually invented as an idea back in the 1930s and 1940s and so there's been this basically this 80-year Arc of thinking and in fact the original neural network paper that uh you know Chad jpt and all these other systems are based on was written in 1943 three right so this has been like an 80-year journey and AI has kind of been kind of alongside the computer industry and the internet you know the entire time and you know some companies have figured out certain things to do but it's never been the the kind of main thing in the industry um and then of course there's been this huge breakthrough in the last five years where all of a sudden it started to work and so it it it it feels like the the moment has arrived um and maybe the you know there's a lot of interesting things to talk about but one of the most interesting things is why is it happening now um and that goes you know a lot to the topic of this conference is a lot of why it's happening now is because of data it turned out one of the things you need to make AI work is just an enormous amount of data and so it actually turned out we had to get the internet to scale right we had to get like the full Corpus of the world wide web and you know basically the full crawl you know that goes into the search engines and we had to get um you know all the image data on Google images and videos and so forth uh to be able to train these things and then it turns out they actually work and then of course that means to make AI work even better now means we need to feed it a lot more data um and so it feels like these world of like internet data and AI are kind of all slamming together um and and and now magic is happening that's amazing actually you had this argument that I read somewhere which is kind of interesting is that the previous kind of generation of things you know feed the Next Generation well can you explain that you know you said you know books reappear and oh yeah yeah so so there's this interesting thing so Marshall mcluhan um was this famous media theorist uh 40 50 years ago and he had this thing he said um uh each medium becomes the content for the next medium um and so he said basically when like radio uh you know emerges basically what were they doing you know they were basically like reading out newspaper articles uh when television emerged what do they do they basically televise presentations of stage plays uh you know when the internet emerged what did it do all of a sudden it's a platform for you know all previous media forms right including you know all of TV and movies and everything else um and then and then AI is kind of the ultimate example of that which is then all of these different forms of media basically you know sir you know basically uh surge in uh and become components of training AI right one of the big breakthroughs in AI right now is this idea of multimodal AI right so you basically like if you use chat DPT today it's trained on text if you use mid Journey as trained on images but the the new AIS that are going to come out are going to be ones that are trained on multiple media types all at the same time right so you have AI That's trained on text and images and video right and structured data and documents and mathematical equations and so forth and then the AIS will be able to span across all those um and so it yeah they they all basically roll forward and it turns out they're all important so then taking this analogy to the next level what's going to happen next all these AIS are going to feed the next thing that comes isn't it yeah well bigger AI yeah bigger AI bigger and better yeah well as you know like a lot of research in AI right now is basically but so today you train AI with human created data and then you have humans do what's called reinforcement learning where they're basically the humans are basically tuning the result from the AI but a lot of the research happening right now is on how to get AIS to basically teach and train each other right and so there's going to be this like laddering this this sort of upward Cascade uh where the AIS are going to actually they're going to train their their uh their successors so if we get into that then okay so this is so it's fascinating this is going to feed the next thing so the output of these AIS will train the next Ai and actually if their output can feed the next one it's it seems you now get into this recursive Loop but could potentially happen pretty fast let's get into the whole you know uh you know the whole doomsday scenario takeoff so on you said it's a category error yeah uh can you explain that a little bit well so what happens is there's so there's a sort of idea uh recurring idea in human history there's sort of this idea that you're going to have something sort of fundamental change in The Human Experience and then it's going to lead either to Utopia right on earth right and and the Utopia version of AI right is this concept called The Singularity right and this is this this concept originally invented by um this guy for intervene a great science fiction writer in the in the 70s and then Ray Kurzweil has carried this idea forward for the last you know 30 years or so right so the singularity is the idea that you know basically at some point computers get you know basically more sophisticated than human brains and then all of a sudden you get the singularity and all of a sudden everything changes and and becomes wonderful um corresponding of course to the idea of creating Utopia on Earth there's there's also the other idea which is you know creating dystopia right creating hell on Earth um where everything just like goes you know completely to crap um and so the problem I have I'm an engineer by background um and so the these the these sort of very rapidly to me start sounding like science fiction scenarios um so so I don't think this is actually what happens um uh there's this guy uh who I like is writing a lot and he uh up in Berkeley and he he has this term uh he calls what we're doing is a species basically a civilization as he calls we're slouching towards Utopia right and it's I I like that term a lot which is like basically it's like things are getting better on the margin kind of all the time like you know material welfare and and and health and and uh and uh you know intelligence all these things are kind of you know the capabilities of what people can do are kind of getting better all the time but they're not getting better in a way that leads to you know basically you you know an actual literal Utopia but you know but we're slouching towards Utopia like in our in our in our imperfect and broken and fallen way right we we do still manage you know someone to improve the world over time so it's sort of a I would say cautious form of optimism not not the radical form makes sense so what about the people that then say there's this thing called I think instrumental conversion which actually can you explain it so there's this concept of the AI no matter what you know goal you have in life eventually you will Converge on some sort of things that you need to do for instance Gathering energy is important because otherwise you couldn't get towards your goal or you know uh you know survival is important and so on so then you know eventually these AIS will realize if they become really really smart that you know uh they might not be aligned with our interests yeah so there's there's a whole I mean this gets this gets very deep so we're going to spend the rest of the day here discussing this is that right this is this is the next one question I hope people brought their water bottles because they're gonna settle in for the next eight hours just on this topic so um Mark told me before we got on stage like if I don't like one of your questions I'm just going to leave I'm just gonna like stop stop right off stage um so um there's sort of these two lines of argument one is like the AI is going to like declare itself its own goals it's like a Terminator scenario where it's gonna wake up one day and decide it hates us yeah uh and my answer to that is like there's no it like they don't they don't it's not a person like it's not a it's not a it doesn't have like it doesn't have Consciousness it doesn't have like a will uh it doesn't have any of these things and then they have this other argument that's sort of so-called AI doomers uh projecting sort of this dystopia this other view of like well it actually doesn't need to ever have conscience Consciousness or sentience or any sort of will of its own it's just gonna somehow like despite it's basically gonna like they're gonna like hyper optimize for these small goals um and paper clip yeah so this famous paper clip Optimizer like yo somebody will make the mistake of telling an AI to like make paper clips and it will basically decide that it needs to convert all atoms on Earth to paper clips right including it needs to like Harvest all human bodies internal atoms into paper clips um and so doing you know in order to like maximize the number of paper clips in the world it will develop its own energies you know it'll figure out how to develop nuclear fusion and it'll have like its own space stations and it'll have like you know its own robot armies it'll have whatever it needs to like maximize the number of paper ups and it's just like I don't know like like okay like you know is the problem that it has its own will or that it doesn't have its own will right and so I I kind of trip it up there and then there's just like practical limitations right and so like a practical limitation is okay where is it going to get the chips to run the sophisticated algorithms to make all the paper clips because like sitting here today we can't even get chips to like run AI at our startups so I have these like fantasies that there's like a baby evil AI you know probably running in the lab at databricks right now with with the Mosaic acquisition it's this like evil evil baby AI that like wants to take over the world but like literally it's it's sent its purchase ordered Nvidia like literally it's just not getting the chips uh to be able to do it so I don't know I think we should I think we should wait to see I think we should wait to see the first evil baby AI before we get too concerned about the big ones okay so let's flip it over so uh you know you said it actually it will save the world right and you listed a whole number of things that it has the potential to make better which one is your favorite and can you walk us through some of those that you're the most excited about so the reason to be very positive on AI is because it's it's in this concept of intelligence um and intelligence is one of these things it's a very very interesting topic in and of itself but we actually know a lot about intelligence because we know a lot about human intelligence and human intelligence has been the biggest topic of study in the social scientists for the last uh for the last century and there have been many thousands of research studies done in the implications of having higher levels of intelligence in people and it turns out basically when it comes to human beings intelligence makes everything better right and that's a very big statement but it turns out there's like a ton of a ton of research to to uh to support it so intelligence basically means that people are going to be more people people with higher intelligence are more successful academically they have more successful careers their children are more successful they're healthier they live longer they are also less violent they're better able to deal with conflict um they are better able to solve tricky problems they're also by the way less biased less bigoted you know they're more open-minded they're more open to new ideas um and so basically like intelligence uh apply to humanity is like this thing that basically makes everything better and then we look at the world around us right including being able to be in a you know in a space like this and with with with each other and all the you know sort of work that goes in you know creating you know the the this kind of environment and everything else that we do um and it's all the result it's you know it's not the you know if we didn't wake up one morning and all these wonderful buildings and electricity and everything was just like sitting here waiting for us as human beings like it got built by human beings through the application of intelligence um and so so we've built everything we have that we like that makes the world work with intelligence we've just always been limited by our own right our our own capabilities our own organic capabilities and now we have the opportunity to apply machine intelligence to all of these efforts um and and and basically do an upgrade of everybody's capability to be able to do things in the world um you know personally the the part that I find most exciting is I have an eight-year-old and the you know the thing that's sort of most kind of emotional for me is the idea that kid you know every kid from here on out mine and everybody else's kids from here on out will grow up with an AI teacher coach tutor Mentor advisor right Ally right that will basically be with you know him or her for you know for for their entire lives um and will be doing everything possible to try to make sure that every person you know has basically reaches their full potential um I I rolled out uh chat GPT to my eight-year-old about about a month ago uh installed on his uh you know put on his laptop and um I was I was you know and I and I was surprised like this is the most amazing thing I've ever done as a parent for my kid like I feel like I'm bringing like fire down from the mountain right uh you know to like buying a computer to my Offspring right it's like I'm bringing him like the most important thing he's ever going to have is like the ability to access this thing and I and I load it up for him and I'm like look you can ask it a question and it will answer any question you have um and of course he's like okay yeah like like duh like what else would you do with a computer right like obviously it's going to answer all your questions and I'm like no like this is really new this is really exciting and he's like okay Dad um so notwithstanding the fact that he didn't understand the significance of that I understand the significance of that um and and I think I I think kids I think kids are going to grow up in a very different much better world uh we're like we were programmers so we were to punch cards for these kids right I remember every single step that the computer industry had to take to get to the point where you can like answer any question and like yeah no for him it's just obvious so is is this the new computer I mean some people have argued this is you know the phone Neumann architecture memory CPU instructions programming with you know precise semantics that was the previous computer we re-implemented in this old computer this new computer that you know prompt engineering and you know knowing how to talk to it and get it to get it extract information you need from it this is a new your computer and that's what people should learn do you think that there's some truth to that or not really yeah so there's a whole there's a whole backstory here so there's this great book called rise of the Machines where it goes through the history of basically the history of AI which in in the in the early decades back in the 30s 40s 50s was was called cybernetics and um it turns out they actually had John Von Neumann actually had there was argument at that time that that he and Alan Turing and all these other guys had back in even the 1930s even before they got to the point where they had the electronic computer um and the argument basically was should they knew they were going to build the computer they knew this thing was going to happen right and this this is it was an old idea you know babich had the the different engine idea and so there was like a seven year eighty year period where they were thinking about how to build you know they had calculating machines and tabulating machines and these things so they knew they were going to build general purpose computers and they knew that technology was getting to the point where they could do that and they actually had this big argument back then should the computer be basically should it be an evolution of basically a calculator right so should it be a thing you know now known as a Von Neumann machine should it be a thing that basically executes deterministically right sequences of instruct actions that are literally exactly what the programmer tells it to do right or should you basically not do that at all and should the computer be based on basically the model of the human brain and right they actually knew about neurons like they actually knew quite a bit about about neurophysiology at that point so they said the neural network paper came out in 1943. um and there were quite a few people at the time who argued no we should not we should skip the Van Norman machine we should just go straight for the brain but but they did right they didn't have the CH you know they didn't have the chips they didn't have the data they didn't have like all the all the all the underlying technology so they couldn't do that back then but but it was even obvious back then that there were these two different two different vectors to build two different kinds of computers um the way I think about it is the the computer we've had for the last 80 years the Von Neumann machine is basically super linear super deterministic right super like like every programmer knows like if if uh if you program a Von Neumann machine computer um and it makes a mistake it's always the programmer's fault right 100 of the time is the programmer's fault it's like you know so programmer you always feel guilty all the time because you're like you know how many bugs are in your code um this is a new kind new computer this is a I call it I call it neural networks basically they're probabilistic computers right and and what does that mean which is like you ask it the same question twice it will give you different answers right and so just the basic idea that you have a computer that you ask if the same question twice will give you a different answer it's just like an amazing thing um if you word the question different ways it will give you different answers right um uh you know if you I don't know catch it on a bad day it will give you different answers like if the training data changes a little bit it will give you different answers um if you stroke it like if you complement it right or you like you tell it to like model its answer after like certain famous people or you know you you do all these basically these things so called prompt engineering is sort of this form of sorcery it actually gives you different answers and then it does this amazing thing which is it will hallucinate right it will literally like if it doesn't know the answer it will make up the answer and people look at that and they're you know sort of the engineering mind you're kind of horrified by that but of course you know if you're if you have a creative kind of bat you look at and you're like oh my God the computer is actually making stuff up like we actually have a computer that's like creating fiction which is like a fairly amazing it's like us basically yeah so basically it's a lot like us well this is my conversation having you know a lot a lot of friend a friend of friends of mine are like yeah you know I you know I don't know whether I can use this thing because I don't know whether the answer is correct and I'm like well have you ever worked with a human being right like this is what we're like right and it's like yes if a human being tells you XYZ it's like yeah you might at some point want to double check to make sure that they remembered everything correctly right but the reason you talk to other human beings because other people have ideas that you don't have and they they you know they they create ideas that you don't have um and and so you know it's it's amazing we now have both of these and of course what's going to happen is they're going to get integrated like you're going to end up basically with hybrid systems you already have this with chat jpt you can you know if you just ask chat GPT math or science questions it will often get them wrong but if you use the Wolfram Alpha plug-in along with chat GPT all of a sudden it starts to get all that stuff right and so I I think a lot of what's going to happen now is sort of a practical form of engineering well you'll bring together both sides you know both of these architectures and you'll have computers that can both create and also you know answer you know do things literally so what does it mean for us like I mean you know come on many of us grew up programming loving it I mean it was the art form that we it was our craft we loved it right we we can't not do it for you know a week of our lives uh what does it mean for our profession going forward the the is programming well I'm not asking if the jobs are going away I'm asking if programming as we know it this very structured you know semantically unambiguous way of writing code is that not really going to be a thing in five ten years and you know we really you're doing this English prompt engineering to you know work with the statistical computer or no programming will probably be around and we'll still need lots of lots of people who can write very precise code right so I think what's going to happen is I think the Java Pro I think most programmer jobs um most of the activity I think it's going to basically go up a level right and so in more and more being a programmer is going to be like being a manager of programmers as opposed to writing all the code yourself right right we're all managers now we're all managers well except of course the the what you're managing are you managing the AIS right um and so and you know the way this can happen is you start you start today you use something like GitHub co-pilot today and it's like okay I'm a programmer I've got the AI and it's kind of helping me at suggesting things you know it's kind of you know fixing bugs for me um and so forth um and and then basically what's going to happen is as as those systems get more sophisticated you'll be able as a programmer to basically give them like more complex assignments and you'll be able to say okay right you know write this code write that code do this do that and it'll kind to go off and do it and kind of report back and and I I suspect what will happen you know and today what you're today it's like a pairing it's like a a person with like an AI copilot I suspect what's going to happen is it's going to be like a person with like more than one of these co-pilots right and so I think you're going to have basically like you know maybe start with two and then basically five and ten and then maybe like really skilled programmers are going to have like a thousand of these AI assistants right um and then um as a consequence of that um you know you're basically going to be able to oversee effectively an army uh of AIS and then it's just a question basically of how much time and attention and effort um you know can you put into kind of kind of overseeing this whole thing um I tend to think the really good programmers are still going to be trained for a long time in the fundamentals of how this stuff you know actually works you know in the same way that really good mathematicians are trained you know you know really good mathematicians are still trained and yeah it's math going away because metal calculators right exactly right exactly right um and so the really good programmers are still going to understand everything all the way down all the way down down to the metal but um the really good programmers are going to be much more productive right they're gonna be able to do a lot more in their careers than they could do before um and then and then a lot of people who can't code are going to be able to actually you know effectively program right and there's been you know long-running Trend through kind of these kind of low code no code tools to get to the point where Ordinary People can write programs without you know having a computer science degree and I think that's really going to go on steroids and so I think a lot of people who are not trained programmers are going to be actually creating code yeah so that's about programmers if we like broaden it to the job market as a whole uh you know some people say hey there's a fixed set of amount of work that needs to be done and we have this much labor to do that work and now we have computers that can do a lot of it so now there's less work to be done so there'll be less jobs right uh what's your thoughts on that yeah so that's sort of there's a classic fallacy in economics called the lump of Labor fallacy that you're alluding to which is right it's it's a zero-sum view of the world where there's a certain amount of work to be done if machines do the work then humans don't have anything to do what ends up happening is actually is actually pretty much the opposite um which is basically when machines can take over things from people what you actually do is you free people up to actually do more valuable things right and so there was a time when like you know 99 of us were basically Farmers right and then later in the industrial revolution there was a time when like 99 of us worked in factories um and sitting here today there's a much smaller percentage of people who work on farms and factories but there are many more jobs in total right because a lot a lot of new demand got created and a lot of new businesses and industries got created and so I I think there's going to be just tremendous economic growth that follows from those huge amounts of job growth wage growth also you know flows flows naturally from this specifically with code you know code has this property where basically the world can never get enough of it like there are always more programs to write there are always more things that you want code to do you know nobody's lacking everybody everybody knows you you know everybody's is in a business knows this like nobody ever runs out of ideas for what they want software to do what they run out of is the is the is the time and the resources to actually build the software that they want um and so I I suspect what happens is both a massive increase in basically the amount of software in the world that comes out of this and then also ultimately a very big increase in the number of people actually working on software that's awesome okay so final final topic I want to touch on because I think it's like really relevant for this crowd here we've been saying data and AI is the future I've been saying that the crowd here you are the future you talked about you know Legends and heroes can you tell us a little bit about that and you know how's your day and actually some advice maybe to the audience where you know we have young folks that are beginning of their career they're thinking about you know they're here Mark and recent you've seen you've been able to see 10 20 years ahead of time many many times before um how should I think about their careers and what should they get into what should they get good at uh looking forward yeah so in my piece so I use the terms Legends and heroes so the Legends I refer to is you literally had people you had literally had people scientists this is sort of the science side of things you had computer scientists mathematicians starting in the 1940s who basically worked for 80 years without really getting a AI scientist who without really getting a payoff for their work and I just remember I was when I was in I was in college 89 to 94. and I remember you know studying AI at the time for for computer science and AI at that point was like a Backwater like it was like a discredited it there had been an AI boom in the 80s that didn't work like it was a bubble and a burst and it was a bust it was very bad at the time and so by the late 80s AI had been very discredited and this had been about the fourth of those Cycles where AI basically there was a lot of promise and then it just didn't work um and so I remember that you had you had scientists working on AI and computer science departments and in in Labs you know where they would literally like be born like grow up like get their phds become professors teach AI for 30 years retire and have nothing to show for it right and that's the way I mean they're not even pass away like they're yeah exactly like they a lot of them are not here today and you know literally like they were working on a set of ideas that you know we now know like those ideas were going to work but it it took 80 years and so the the level of determination and vision right and courage and insight and stubbornness like required to like devote your life to a field where you never get the payoff you know I now that we know that if it never works you start to wonder if they were you know completely sane um now that we know that it works you're like wow they saw the future like they really understood it the whole time they you know they got it you know they were they they it just needed time you know for the whole thing to pay off and so they they I put them in the Legends categories and they looked in the heroes category um are all the people who are making it work today um it could be all of the crowds here which is a whole crowd here and and I use that word specifically because I talked about this in the essay like that we're in this we're in this kind of cultural moment where just like people are really mad about like everything like people I don't know people have noticed but like a lot of people are in a bad mood these days um about like a lot of stuff um it's been a while yeah we we the the the country in the world is in kind of an emotional Funk of some kind um and as a consequence of that you know any new thing that happens immediately the sort of arguments come out for why it's bad and horrible and it's going to destroy the world and it's going to destroy it's gonna like be like this this like horrible you know just read the newspaper accounts what's happening it's just like this would be this horrible thing um and so I think anybody actually working to Make a Better World possible through through this technology which is I think what they're doing um you know I I think it's a hero so this is you know so basically people working hundreds of years you can finally reap those benefits so we're really lucky to be alive right now and you know you can all become the heroes of the future yeah that's Mark Andreessen thank you so much everyone thank you thank you everybody awesome thank you man thank you [Music] thank you Mark and Ali for such a wonderful Fireside discussion today we're going to hear from the creators of these incredible projects that you see listed up on the slide here we're going to kick it off with the data projects starting with Apache Spark what started as a research project at UC Berkeley Apache spark has become the de facto big data processing framework out there I love the collaboration across industry Academia and open source to truly Embrace and contribute to this project for our next speaker we're going to have the co-founder and Chief Architect of databricks who still manages to find the time to be one of the top overall contributors to Apache spark to tell us about the latest and greatest in Apache spark Reynold chin please join me in welcoming Reynold to the stage good morning again so as I already told you this is the 10th year of Apache spark um 10 years ago we at UC Berkeley donated the Project Spark to the Apache software foundation so spark officially became Apache Spark in the 10th year I would like to take a little bit of opportunity to review yourself the Project's history seemed like the theme I'm doing on this Kino stage this year anyway so smart start at UC Berkeley the first place of not just spark but also BSD Unix postgres risk rate there's a lot of foundational projects right one thing that's really special about the project is that it has a culture of incorporating bleeding edge research ideas and make it real make into the real world workloads and many of this led to also high profile research publication really moved the field forward and this really what make the spark project very unique in its beginning but the project have kept its roots couple years after the donation of the project it has received some widespread media attention for example IBM called spark potentially the most significant open source project for the whole next decade I think that to a large extent become true now you might ask hey who cares about what IBM thinks I'm not using mainframes anymore in the same year Fortune magazine called spark the Taylor Swift of data it's not the first time we showed it at the conference I finally felt hey when I read it I thought hey I could finally explain to my wife and my mom what I do for a living the uh what he means is really spark came from nowhere with the meteorotic rice and Rose to prominence and ubiquity all right according to OSS insights which is the open source project a project that tracks all the open source projects for the every year in the last 10 years Spar has been the most actively developed project in Big Data with over 3600 contributors and forced out forty thousand commits is remarkable for a 10-year history now but perhaps the most remarkable remarkable metrics will ADI show you uh yesterday the project has been downloaded over a billion times a billion times that's across Maven pipei and also on databricks itself so what other reasons led to the Project's ubiquity I think there are a few principles behind the project really worth highlighting the first is that the focus on simple expressive and modular apis with well-defined semantics it makes the program easier to write but also more importantly this API sufficiently abstracted so they allow the backhand to optimize over time without the programmers having to actually change the code they write the second which is often overlooked is spark can run virtually everywhere you can start developing your spark programs on your laptop without any internet connectivity I was literally coding the last time I was on the airplane spark programs running on my laptop and you can use your cicd tools in your own network environment and you can publish your spark program to the private cloud or the public Cloud for execution this is very important because you don't have to depend on any third party everything is self-contained you don't have to handle any third party just to even test your program now last but not least the multi-paradigm extensibility of the Project Spark is available there's many facets to that Sparks available in virtually so all the most important popular programming languages and data and AI started with Scala we added Java R python NCC coordinator but there are other ways to get a stand Spark for example data sources and Federation it's been extended to virtually all data sources any data source you name you could probably find the open source implementation of that data source for spark out there and we're in this generation AI conference here spark is a very important part of data science and AI the four most popular AI Frameworks right now are probably launching hugging phase Pi torsion XG boost and all four have native spark Integrations so despite success the Project's not sitting idle right we're introducing the community is working a massive number of improvements literally thousands of them for Every Spark release we won't be able to go through all of them but I want to highlight three today that's my personal favorite the first Spark connect and then Python and then a new programming language so we'll first start with spark connect last year on the stage Martin and I introduced spark connect to you it's a new way of invest spark in your applications going Beyond just SQL this year I'm excited to tell you that spark connect is GA in spark 3.4 now it creates a narrow waste for Apache spark computation and this narrow waste can be leveraged to create thin clients that can be embedded into different programming languages applications and Edge devices so what are the use cases for Smart Connect first you can use spark connect to connect to a remote spark deployment in the cloud in your private data centers or in a different laptop you can interactive develop and debug in your Ides sharing the same production environment of your actual environment enables also developers around the world to much easily create newest programming language sdks for spark for example I already know their community members working on the go language support any go programmers here the Scala 3 rust Swift and also our own databricks connect has been upgraded to build on spark connect the second thing I want to cover is python as most of you know python is the number one programming language in the data and AI era but beyond that python actually just ended C and Java's 20 year round in the Tob index we started years ago projects then to make python the first class citizen for spark to really make the spark pie spark experience pythonic and there's been a lot of work that got into it but perhaps the most evident visual one is the autocomplete in spark 2 which is before project exam was started if you start using a python notebook or Jupiter notebook IPython kernel you'll barely see anything in autocomplete for spark apis in spark 31 you'll start seeing parameters names showing up so it makes it already easier to use in spark 334 the latest release you can see full signature along with the full documentation in autocomplete directly making it much easier to write your spark programs like now maybe you don't need autocomplete anymore very soon but I'll get to that so a lot of them are focused on projects then in the last few years has been on the end users experience of writing python code using spark now the community is actually discussing how do we make that first classes and experience of python also extend to extending spark itself in the past for a python shop you want to extend spark to add your own data sources to define a uh you should find table function you have to learn Scala or Java because that's the only way to extend it now the community is discussing creating python-based data source apis and user table to Define table functions so python teams can extend spark without learning any new languages and the committee is also working on how do you properly test Spark It's a remarkably complicated process to test any data programs it depends on not just the program's logic itself but also the data so we're creating a new library that introduced native research for data and a lot of other capabilities make it substantially easier to test your Pi spark program with all of this work we're really making python a first class citizen going all the way from the end users experience to the experience of framework developers and more sophisticated users that can actually extend spark itself so the last one I'm talking about is the new program language some of you might have guessed but it's actually a little bit different from what you think about the so by now you all heard Andre capacity said the uh um the hardest new program language is English my extension of it is that English new program language generally as a new compiler and Python's a new bytecode what's that mean in the last few months a lot of you probably have tried using tragedy to generate spark code and not because you didn't know how to write spark but because sometimes it's difficult to figure out exactly what API to use or what signature you need the however tragedy video trained on the very large corpers of Sparkle on GitHub and probably on open internet that's full of anti-patterns and full of patterns that might be owed um so he actually also often generate code that's not ideal and in order to improve the code you start using prime engineering to tell chat gbt hey please do for example use the data frame API don't use the rdd API and a lot of you are doing that every day so instead of you having to figure out what the right prompts are we're excited to introduce what we call the English SDA SDK for spark it's a new open source project they help you author spark code with prom engineering already done by the spark experts to minimize anti-patterns inside me walking through you through it I would like to just show you what you can do so with that I would like to invite Allison onto stage to show you a demo [Applause] [Music] thank you Reynold just set up the computer here now let's do a demo for this demo why do I do some data analysis on spark I want to see the community contributions over time using the GitHub pull request data we can use the GitHub apis to download the data but wouldn't it be great to use spark to read the data directly with the latest proposal to add python-based data source apis we can write a custom GitHub data source in Python and use it to load data into spark as a data frame now we have the GitHub data let's do some data cleaning and transformations I want to get the seven day moving average of the number of requests created by date I know I need to use some aggregation and window functions to compute the moving average but I don't recall the specific apis I should look it up so here's window function documentations and stack overflow what we can see here is that I can already Express what I want to do in English but why does it take so much time to put my thoughts into code can we just use plain English absolutely we're able to do this using the English SDK for Apache Spark it is super easy to get started you can simply instantiate the spark AI class and activate it and all your data frames will be AI powered let's reload the GitHub data and do some data Transformations using English first let's add a column called date it's copy pasting here then let's aggregate the number of pull requests by date yeah just copy paste in the prompt here and finally we can compute the seven day moving average copy this over and now let's display the results to see if it works great we have the data now let's visualize it to see what it looks like the Flawless good but wouldn't it be great to also overlay the historical spark releases on this data and check for patterns instead of manually copy pasting the release data I can request the data through the create DF function and filter for major and feature releases let's see what the data looks like finally here is what I want to plot we have the seven day moving average and the vertical bars indicate the spark releases everyone who has ever tried to plot something like this knows how painful it can be but now using the English SDK it's much easier you can use it to plot the data with a prompt here I'm simply typing out the plot we want to generate seven day moving average annotated with spark versions let's run the command and the results look exactly as we wanted look how easy it is balances [Applause] the number of people that could create this chart without looking up the internet spending half an hour probably in the hundreds in the entire world right but now all of you could do that very easily using this uh Pi spark and by the way it also looks pretty obvious that everybody procrastinates it's no longer on the screen but they look at it right before the commit deadline the version release deadline the commit activity Peaks so spot committers totally procrastinates and right after everybody takes vacation in the uh the commit drops massively so the English SDA for Apache spark does not just help you write your basic spark code it actually helps you with all the stages of your data science or data engineering program going from ingestion transformation verification explanation to applauding all right go check it out at Pi spark.ai for the English SDK for Apache Spark now there's so much more that's happening with this Apache spot community that I simply don't have time to talk to you all about and when culture encouraged you to check out this field talks about it at this conference today now next I would like to invite Michael ambros on the stage to tell you more about Delta 3.0 which is probably the most exciting release of the project since the initial introduction of the project Michael needs no introduction he's a distinguished engineer at databricks one of our earliest members of the company he's also the original creator of spark SQL structure streaming Delta Lake Delta live tables list goes on all right welcome to on the stage Michael [Music] thank you Reynold it has been an exciting year for Delta things are only continuing to accelerate in the last year we've gone up to processing two exabytes of data per day I remember the first time I heard this term it was just an unfathomable amount of data in grad school that we heard Google was processing and together with you we've now scaled Delta to do this twice every single day Delta is fast it processes in some ingestion use cases more than 40 million events per second and it's popular we had over 500 million downloads in the last year one of the reasons it's so popular is because it's reliable it's been battle tested by more than 10 000 companies in production it's open with over 500 contributors adding into the ecosystem and we're not slowing down we've got over 80 features that we added in the last year and I'm going to talk about some of them today one of the really cool things about Delta is it's the fastest lake house format and this was recently validated by a study done by UC Berkeley and Stanford and also validated by an independent third party what you're seeing here is the performance of Delta on tpcds and you can see we're faster not only at loading data but then also querying the data once it's been loaded into a table one of my favorite Parts about the Delta Community is Delta RS for those of you who don't know about it Delta RS is a complete from scratch implementation of the Delta protocol written in Rust native code and it has exploded in the last year this is completely driven by the community and it's really taking off right now the pretty cool thing about this is because it's in native code you can use it to link into other ecosystems like Python and this is you know lots of people are starting to use it this way you no longer need a whole jvm just to query Delta tables another unique thing about Delta is Delta sharing it is the only open protocol for sharing massive amounts of data without being locked into a single vendor's compute platform the way it works is we actually sign individual parquet files on demand so you're still securely sharing your data but you're always sharing the latest copy and you don't have to make extra stale copies of it the ecosystem has been exploding with over 6 000 active consumers and over 300 petabytes being shared per day on Delta sharing and as with any sharing system the most important part is the ecosystem and as you can see there are lots of people sharing data on Delta and I'd like to focus on some of the newest that we've just added recently so Oracle probably the most prolific relational database ever now supports Delta sharing so if you want to do analytics on that data stored in your transactional Warehouse you can do it without creating expensive stale copies and similarly twilio is now unlocking all of the Power of their segment customer data platform so that customers can natively query it with any engine that supports Delta sharing there are a bunch of new features that we've added including support for structured streaming as you can imagine when you have a huge data set it just doesn't make sense to scan the whole data set over and over and over again you want an efficient protocol that tells you only what's changed since the last time you read it and so now it is deeply integrated with structured streaming we've done tons of improvements on the back end and so now we've improved query latency by up to 50x we have full support for oauth 2 spark and pandas and then finally for the CFOs in the audience there's a pretty exciting development when you shared Delta sharing data on cloudflare you can now do that with zero egress fees which is pretty great if you know how expensive Cloud egress fees usually are last Summit we announced Delta 2.0 and we haven't slowed down this year with Delta 3.0 we have a whole slew of new features and I want to focus on three of my favorite the Delta kernel uniform and liquid clustering let's start with liquid clustering so liquid clustering is solving the problem of rigid partitioning which is often required to get good performance but is very difficult to get correct let me start by explaining why it's so difficult to get correct even if you pick a pretty reasonable partitioning partitioning by date and by customer there's going to be skew in your data and enforcing that every single file has exactly one distinct set of partition values is a pretty rigid way to do this and so you can imagine cases where a large customer has way too many files packed into a single partition and you just can't filter beneath that granularity in contrast small customers will be forced to create lots and lots of tiny files which is both going to blow up your cost and blow up through your query times when you're querying this data liquid clustering breaks this down so that we no longer have to partition at these rigid boundaries instead we can look at these smaller data sets and collapse them across this dimension we can also do it across the other dimension we can even do it in both dimensions at the same time picking the optimal file sizes based on the actual data that is present in your table and it works in both directions it's not only about combining data it's also about splitting those whales up so you can filter at a finer granularity the best part about liquid partitioning is even though it's less tuning it's also faster so when you compare it to Hive style partitioning where our Benchmark didn't even complete you can compare it to Z order clustering which was the old way that Delta did multi-dimensional data clustering liquid partitioning is up to 2.5 times faster and it's not only queries that are faster it's the process of ingestion that is faster as well when you compare it with Z order there is dramatically less right amplification because we are gradually clustering the data as it's added to the table rather than only in spurts when you run the optimize command one of the best parts about Delta is its ecosystem and I want to highlight a couple of the new connectors that we've just added we now have support for Flink which I bet you never thought you'd saw at a smart conference we have support for trino and the whole python ecosystem but one of the problems is at the core of this ecosystem is the Delta protocol and we're continuing to evolve this protocol and we want to avoid getting into the state that we've seen happen with some other lake house formats whereas they add new features vendors won't implement the complete spec and now you're forced to choose between using the latest features or querying your table wherever you want to and that's why we're really excited to announce the Delta kernel it's basically going to take this eight different implementations of the protocol spec that exists today and collapse them down into one unified group we have a jvm ecosystem we have a native code ecosystem and it handles the complete Delta spec both the metadata and the data so you don't have to worry about deletion vectors or column mapping or whatever else we come up with next in the Delta protocol you just get your data back from the library now as much work as we do on the ecosystem there will always be other engines and we actually think that's great it turns out that any lake house format is better than a proprietary system but since we can't control which formats these other engines are going to be able to query this is a big problem for some of you deciders in the audience you don't want to pick the wrong format today and then find out tomorrow that it is forcing you to use some proprietary system that doesn't support Delta well you have to convert all of your data at that point that's a very scary Prospect but if you squint a little bit it turns out that's actually probably not necessary if you look under the covers all three lake house formats are based on the same fundamental principle which is an age-old technique in databases multiversion concurrency control basically all the different systems store parquet files when they make changes rather than update the parquet file in place they make a copy of that file and then the real transactional magic comes from this extra metadata that sits on top and says for the this current version of the table which parquet file should I read and since all systems are based on the same underlying principle we're really excited to announce Delta uniform what Delta uniform does is we have one single set of parquet files and then we can create the metadata of any lake house format so you can query your Delta table anywhere in the open lake house ecosystem so we're going to take this fragmented ecosystem and bring it all into Delta and you might ask yourself is it fast it sounds really expensive to create all of these different you know copies of the metadata and it turns out as I said before the metadata is actually a pretty small part of it it actually costs less than five percent on your right times to enable extra formats and what's even more exciting is it turns out Delta is better at writing Iceberg tables than iceberg is when you use Delta to produce an iceberg table yeah that's like yeah because of the advanced clustering I was just talking about when you use Delta to produce that Iceberg table you have better parquet files and thus your queries run faster so this all sounds a little bit too good to be true and it wouldn't be a spark Summit if we didn't do a demo so I'm going to move over to my laptop and let's see how this works it's a big stage foreign okay so here we are in a databricks notebook and you can see a very standard create Delta table statement and all I'm adding to it is this one extra table Property Delta Universal format enabled formats and I can just list which other formats we want we're releasing with full support for Iceberg and we're working with the community to add support for hoodie so I'm going to go ahead and run this command and then I'm going to switch over to another cloud data warehouse and as you can see we've got our table here and if we look at it Google bigquery thinks this is an iceberg table so let's go ahead and create a new query on it go ahead and add a star here and I will click run and there it is it's actually a Delta table pretty cool [Applause] I wanted to show at least one other cloud data warehouse but unfortunately their open format support is still in private preview and I couldn't get access but if you do have access I encourage you to check this out and with that if you'd like to learn more there's a bunch of other sessions today you can learn all about the exactly how uniform is implemented in this session on Iceberg and Hoodie in Delta Lake you can hear one of my favorite use cases where they Adobe actually converted from uh Iceberg to Delta and you can also learn how to build your own Delta connectors with the Delta kernel thank you so much [Applause] please welcome to the stage original creator of Apache spark and ml flow Chief technologist matte zaharia [Music] okay hey everyone um very excited to be here to host our next event which is an AI faculty panel with faculty from some of the top AI groups in the world so we've heard a lot from folks in business uh from customers about AI but I also want us to hear a little bit about what's happening in AI researcher and you know what these researchers who are leading the way are excited about next so I'm really excited to introduce three members of the panel today we have Mike carbon who is a professor at MIT and also a co-founder of Mosaic ml when we started the panel you know it was before anything was even happening with Mosaic ML and databricks so awesome to have them here we also have Don song from UC Berkeley who is an expert in AI as well as security and Daniel larus who is the director of MIT C cell you know arguably the birthplace of AI research in general so welcome them all to the stage with me [Applause] thank you and huge huge pops for Daniela for coming here despite Carter's being injured she's she's amazing so you know we didn't even know about it until until very recently so uh yeah um so so welcome everyone so let's start with uh with some introductions um if you can just each tell us you know kind of what you work on what what have you been doing you know recently that you're excited about in AI research to give the audience a sense of it and maybe Mike you can begin oh yeah sure so I'm here I guess I'm here with my two hats on so my academic hat where I've been doing research for about 15 years and of course my Mosaic ml hat where I'm a founding advisor and you know having a great time doing work there as well and of course you know potentially our plans to join the databricks family shortly and so if you look back to the history of what I've done where I come from I come from computer systems and particularly my entire interest has always been in I don't make computers go fast how do I make them more efficient and I do a lot of work as well in the programming languages and compilation space so how do we look at the structure and organization of programs again to make them faster and more efficient press for this audience out here in the data space and the database space thinking about things like query optimizers that's where I spent a lot of my time but you know how did I get into AI well and some of the work that I've done with my students over the past several years what we've done is we looked at the structure and organizations of neural networks themselves I notify there's parts that perhaps don't matter so we can take make these very large models turn them into much smaller models perhaps an order of magnitude smaller but still train just as well and still give you good amazing results and over at mosaic ml these perspectives towards efficiency is what we've been driving based on those insights of how can we take these large models and make them much more accessible for for you and everyone else that's out there right these models aren't only in the hands of the largest organizations they can be in everyone's hands great yeah Don great yes um so I'm a professor increase Berkeley and I also have a startup working at the intersection of AI security and decentralization as well so um there are so many exciting you know areas that we have been working in the areas related to AI so here I'll just talk briefly about the areas at the intersection of AI and security so given all the amazing advancements of AI capabilities LMS and so on so of course one application domain that's really exciting is in the security domain all these advancements in LMS and so on can really help a lot in the Security application domain for example some of our recent work show that by even just using you know GPT and these large language models actually it can really help significantly improve for example security adults of application for example we were able to show that it can significantly improve the manual audit efforts for security for smart contracts and also we show that by using these large language models change um for example blockchain transactions and their execution choices we can identify anomalous and malicious transactions and for a broad spectrum of different applications that we examines these trained models can actually Rank and attack transactions in the real world to be actually the most anomalous transactions for over a dozens of different real world applications and these attack transactions actually have caused close to 300 million dollars worth of damage in the real world so this is just another example illustrating the power of these new large language models and the AI capabilities and on the other hand of course these type of capabilities also pose a lot of challenges that's also something that we have been studying for example showing that these large language models they have big privacy leakage problems and also some of our recent work in collaboration with the meta has been on how we can enable privacy preserving AI model fairness assessments and this actually is the first half is Kind large-scale real world rollout that's been loaded to Instagram users to help in for the first time in a privacy preserving way to assess the fairness of AI models that's actually being used in the real world so these are some examples of exciting areas that we've been working in very cool and Daniela well so I'm also here with multiple hats I'm a professor at MIT I I run csail my students my former students and I have a number of companies in the space of AI and it's just such an extraordinary moment for for our field and in particular for AI now at MIT what I'm most interested in is is really asking some foundational questions about what AI can achieve and so if we think about the fact that the majority of the successes we hear about today are due to decades-old ideas that are now empowered by data and computation it's kind of natural to ask what else is there because if we do not come up with new ideas in time everyone will be doing the same thing and the contributions will be increasingly incremental and so I'm super excited about a new approach to machine learning which we call Liquid networks it's a compact approach to machine learning for it's a kind of a continuous time approach to up for applications that have time series data liquid networks are compact provably causal and they have very nice generalization properties when trained in one environment and applied in a different environment with really a huge distribution shifts and so um yeah so what the world is trying to make networks bigger and bigger I want to make them smaller and so to give you a sense of what you can do with liquid networks um in our work we do a lot of um Edge device and robot applications and so for instance if you want to get a robot car to stay in lane and steer it takes about a hundred thousand diploma networks to get good behavior and it only takes about 19 of our liquid networks and so if you have a solution that involves only on the order of a couple of uh of of nodes or maybe on the order of tens of nodes you can then extract decision trees to explain how that system makes decisions and in doing so you can begin to address safety critical applications where knowing how the model reaches decisions is very important that's super cool yeah I do think now that we've gotten you know neural networks with gradient descent working people will discover so many other ways of doing you know this kind of computing and in a much more controllable way so super exciting to hear about that um so one question I I had for all of you is um you know as researchers you see a lot of the emerging ideas and Ai and here we have an audience with many practitioners who are you know building things every day so what's an idea or an approach you think is is coming up in UI that um you know in an AI that is looking promising and that you think practitioners need to know more about you know that they aren't paying enough attention to um so curious what each of you think about uh about that if whoever wants to yeah the eyes around me all right uh let's see so I guess so we're at a big data conference I guess I actually want to say small data uh and so what I mean by that is there have been some interesting results um some things that we've been working on some things that other people have been working on uh and a highlight I think you can look at from the past couple weeks so maybe a month old at this point there's this paper textbooks are all you need now it's important if you see a paper that's out there in Academia sometimes it's less about the result and more about the idea as we're talking about here but what they were trying to do they they wanted to go out and build a code model like a code generation model sort of like what you're seeing yesterday in the lake house IQ example the lake house Q demo where large language model is going to generate code and the typical way that you're going to do that now is you can find as much data as you possibly can scrape the internet stack Overflow all these data sets that are large and pretty dirty in some ways throw them into a model and this thing will amazingly generate code really great but these models are large right we're looking at chat GPT you know in that at least in the 3.5 chat gbt with gbt 3.5 and that 150 175 billion parameter range these are large these are massive right and we we think these models are perhaps out of the reach of many but in this paper what they did is they find again small data if you do a very good job of trying to curate your data you can find these opportunities to process on less data orders of magnitude less data and with orders of magnitude smaller models so they had a 1.3 billion parameter model that was besting some of the best models that are out there in the 15 billion parameter range and even exceeding at least in their domain evaluation the results that you're getting out of chat gbt driven by GPT 3.5 and so I think that's the future the future of Focus right Chad gbt is absolutely amazing you know it knows everything in all these different domains but when you can focus on your particular domain in your use case I think they're really good opportunities that for building these models and building these models yourself at a scale is accessible to you and so that's the direction I see because I know everybody's thinking about chat GPT but these new coming directions with these smaller models I think are very promising very cool yeah um I think uh for practitioners in general you know practitioners are really excited about these new capabilities I want to use them in different applications and so on one hand it's really exciting to see all the great progress and what you can cool things you can do about it but on the other hand I think one thing that usually practitioners maybe don't think enough about that's the thing that I wanted to emphasize here is that often because these services are exposed to you know the general users and so on like you have to think about how these Services may be misused for example um these air models with our work and with you know other research work in the community these models have a lot of security privacy challenges and issues and they can be easily you know this kind of jailbreak and also they can be Foods in giving wrong answers and also as practitioners use um models out there actually there has been you know our work and others work have shown you can very easily implant for example this we call them back doors or children's into these models that's very very stealthy you don't even know that they're there but actually um then the models can be triggered under specific conditions to misbehave so this is just the tip of the iceberg there are lots and lots of these different types of security privacy and other types of transporting issues with these air models and how they can be properly used in the application domain in a responsible way and I think one of the really big challenges actually for practitioners for deploying these models is how do you know when the models are actually ready to be deployed how even to measure how the model essentially behaves and how what are the Matrix for measuring the traffic and the different aspects of the model so one of our recent work uh with my you know Wonderful many collaborators is called decoding trust you can actually go to the website decoding trust.github.io to look at more details so this is uh the first comprehensive systematic actual evaluation framework that covers a full broad spectrum of different aspects different metrics in the transportation software models and this is only just the first step we hope that this as open source framework we can encourage the community to come together to further develop this to really help practitioners to know better how their models are doing besides just the performance perspective but actually more in the Sprout trustworthiness perspective to really know when the the model actually is ready to be deployed if not what are the issues and so on very cool yeah maybe so we can talk for a long time about the excitement of of the work in Academia because um despite the fact that we have such extraordinary Potential from AI there are a lot of problems that need to be addressed in May in order to make AI safe and trustworthy and environmentally okay and easy to deploy I guess Don was talking about how do you know when the model is ready thinking about rigorous approaches to testing and evaluation is super important thinking about how we how we build trust in in how we use these how we use AI is super important but if I can pop up at one level I would say that when we when I think about how AI is used today the basically three categories of AI Solutions we have ai solutions that are primarily about pattern recognition and and this is where deep learning and very and kind of fit full and then there are AI solutions that are primarily about deciding what to do and this is the kind of the body of work around reinforcement learning and then there is the the group of AI solutions that are all about generating new things and this is the generative AI space which includes tools like GPT and in each of these areas the Academia is actually working to better understand what kind of solutions we get to better to increase the representation in other words what can these uh what can the models represent and recognize and also to to really understand the properties and I will tell you that in each of these three categories we have issues we have issues around data because they all require a lot of data and um that means that the computation is huge that also means that there is a large environmental footprint I mean did you know that deploying a very small for today's standards model today releases 626 000 pounds of carbon dioxide in the atmosphere this is equivalent to the lifetime emissions of five cars so as we think about deploying AI in the future it's important to keep these numbers in mind because we want tools that support Humanity but also the planet so we have to deal with data we have to deal with computation we also have to understand the black box and ensure trust and privacy in our Solutions and so the good news that is that the Academia is working on all these problems and next year when we come together at this meeting we can report to you on on more achievements makes sense cool yeah I think we have time for a couple more questions so I uh I'll uh so let me uh try to think uh about some of the more interesting ones so okay so one question um I have for everyone uh that I think is on everyone's Minds here is you know what do you think about AI becoming sort of democratized or commoditized so like on the one hand things like like computer vision you can run you know pretty much anything that state of the art on your mobile phone on the other hand we've got large language models which are you know extremely large and it seems making them larger makes them even better what do you think about the technology Trends will this be something that you know um uh kind of everyone that that will get a lot cheaper that everyone can do uh or you know what what are sort of the pro and counter Arguments for that yeah I I've already said it yes I can just say it over and over again yeah I think these are these are the trends that are taking over um and we talk about democratize I I still feel like I want to break that into two different categories uh one is you know we see chat GPS and it's absolutely amazing again like I said you know it knows literature and the Arts Sports pop culture right anything that you can think of right that I mean it's truly truly revolutionary I have to say when I first sat down and started using it I hadn't expected to see these types of capabilities particularly in the interactive code generation sort of getting back to my domain perhaps in my lifetime truly amazing I think in that space we're already seeing democratization in the sense of increased competition right I mean Bard Claude PI from inflection results that we're putting out at mosaic um there's a there's a hot race right um yeah the economics are there to bring down the cost of those models I think there's this other direction as well on can you build these things yourself right perhaps it's great that there are these closed models that are out there that you have access to but could you build this yourself and I think the trends are are positive there as well so going back to what I said before you know this textbooks are all you need results if you can focus right again so cheap can do everything but if you're building a biomedical model do you need your model to give you Sports statistics right or talk about Taylor Swift no right but Which hat gbt you're paying for all of that right you're paying in the terms amount of data small needs to be trained on all this data you're paying in terms of the model size the more data the more recall you need over a larger knowledge base the larger the model needs to be to be able to do that effectively so if you can focus there's opportunity there right and textbooks are all you need there's there's an example there and some work that we've done at mosaic we actually did this in the biomedical setting where we trained a small by again relative measures three billion parameter model on PubMed which is this huge biomedical Corpus that's out there biomedical papers scientific papers and then fine-tune this to actually respond to questions on the USMLE so the U.S medical licensing exam and we set state-of-the-art results of course I was back in December and we were able to beat models that were 120 billion parameters at the time because we were able to focus so for me that it's it's about focus and I think those Trends are in the favor everyone sitting in this room where you've got really interesting data and it's just about Focus being on that data and building malls around that space I can jump in uh if you um so I think that there's already democratization like everyone uses chart GPT no matter what aspect in work and also in life and the tool is extraordinary but um the the concern is that people don't actually understand how it works and so um so this actually causes a lot of problems with um uh with with how the tool is used and how people respond to the fact that now we have this extraordinary tools and um I would say that it's important for the public to really understand what tools we put in their hands so communicating to the public and educating the public is important and that begins with highlighting that um indeed the tools we have today are democratizing our access to information they're kind of giving a microscope into the the Digital Universe but this microscope is going to make mistakes and so it's not about so so using the tools is not about replacing activities but about augmenting the human with some extraordinary new capabilities I like to tell people that they should think about using an AI tool sort of like they think about an assistant the assistant runs around looks for patterns brings interesting data for the decision maker to act on the human really still has to be has to stay in charge the other thing I will say that this demo is causing a lot of fear in the population fear that a lot will be replaced by machines and this is just not the case and so we have to be careful with how we how we excite people about being augmented about being empowered and augmented with new tools rather than being replaced by by these tools oh and yeah one uh one more quick fire round because we're we're running out of time so just you know quick answer uh obviously everyone's very excited about llms now but there are so many other types of data so many other types of applications interactions we can apply AI on uh you know what do you each think is going to be the next day here to get a chat GPT like moment liquid natural Network okay it's very general yeah area area area oh no I'm still processing chat GPS he said short done oh I think I mean also following up a little bit to the pre the previous question and I do think you know this democratization that there are many different aspects to democratization right the negative results in this space yeah right so like this you can look at uh um it's also open versus close source and uh and also ultimate like we talk about this uh personal assistant that helps helps individuals but who actually controls uh this uh personal assistant whether we want to just have the big pet companies have close-ups models to control that's how we want to have open open source models where actually individuals really can have this kind of personal assistance that actually enter their control and really try to act uh in their best interests and following uh or the earlier sets is the reason we have some results talking about the challenges for building open versus Source models um we had this paper on the first prime minister of imitating yeah there's not enough yeah right so like it's not so simple you actually really have to do the hard work to improve the performance and of the base model and so on so there's a lot of challenges and how can the open source Community to come together to make that uh right to really try to close the account yeah so the training process yeah makes sense yeah cool Michael I I think I just want to say multimodal right so yeah I'm required as a programming languages person to say co-generation but um just multimodal so I think we're just in this age where we suddenly have the capability to have multimodal interactions with computers and the fact that that's going to bring so much new capabilities so many people who could previously only access computation via programming right and perhaps those people were excluded I think that's that's just the future more computation for everyone so I do want to jump in because I only had two words um and I will say that I agree with everything my my wonderful colleagues have said but I think we also need to understand what's happening in the box and so um aiming towards causal models is important yeah and the other thing that's important is to get a better understanding of what's Happening Inside the Box because it ultimately we will get to the point where we will need certification or we will need some kind of deployment guarantees when we uh when we put AI in service especially of safety critical applications yeah very cool well thanks so much everyone really awesome to have you here yeah foreign [Music] continuing on with our theme of Open Source projects and democratizing data our next speaker is the co-creator of drb I love duckdb it actually has all of the benefits of a real duck when applied to database management including versatility and resilience but without the Quacks last month alone drb had more than 2 million downloads so please join me in welcoming new ice into the stage [Applause] [Music] thank you Brooke yes hello good morning my name is hannes muleleisen and I'm here today to talk to you about duckdb but I'd like to start with a small story this is my actual car license plate for my first car 20 years ago and what you can probably tell from this is that I may have been a database nerd back then already um I in fact I love these things so much that I went on and do research and this is where I'm at Mark Grassfield who is here today and we were starting to think about new ways of building database systems and we were doing that in Amsterdam at the CWI which probably you have not heard of but it is the Dutch national research lab for mathematics and computer science and it's the place where python was invented by Guido from Rossum all those years ago so it is had had a certain impact on the world but while we were there we started feeling that there was something wrong with the world of data and I should explain it's a movie from the 90s in this movie this character tells another character that he knew there was always something wrong with the world but he couldn't quite put his finger on it and we had the same feelings that there was something sort of off with how people were doing stuff with data and that we should probably do something about it and one thing we noticed after you know many years of thinking about that is that what was happening is that we're people using very big Iron systems like Hadoop to solve fairly small data problems you know stuff that's a bit too big for Excel or pandas or something like that and by using some big iron for something small like here this uh Sledgehammer to Cracker nut apparently that's an expression you just get bad ergonomics in German we have a saying to shoot at sparrows with cannons and in fact there was a development that has sort of supported this where we now have data Lake formats are built on things like Pi K that actually have the ability of turning a huge data set a very unwieldy thing into something much more manageable and we can with formats like okay we can actually very precisely fetch only data that is relevant to a specific query what also has happened is that we got a disconnect between storage and compute and I know you know this and people have talked about this before but I feel it's the difference between you know knowing that something is happening and you know appreciating all the possible consequences of it and one of the consequences that these things are disconnected is that we can scale down compute to the appropriate size that's you know sufficient for the problem and you would be really really really surprised how small you can scale down data problems um in fact the laptop in front of you is more powerful than you think so if the data is small enough you know there's no reason not to put it on your laptop another thing that we've noticed while thinking about you know having the Splinter in our minds about Data Systems was that the community was spending a lot of time thinking about the meat of the Hamburg at a Patty apparently it's called um and this was great you know you can write a lot of interesting research papers about things like join algorithms and things like that but the problem is you're not looking at the whole hamburger right you're ignoring the end-to-end user experience and that was such a problem that people had like aversion towards data system because they were hard to handle and so we decided we were going to work on making an end-to-end good experience for data systems for once so the result of all that thinking and sort of like long 10-year process almost was that a whole new class of Data Systems was required and the first instance of that class is techdb seductibly is an in-process analytical data management system it does SQL and I will tell you more about it in a moment but first the first question we always get is why is it called duct TB it is because I used to have a duck his name was Wilbur he is a very cute duck um duct TB is very flexible it runs really anywhere it can run from a Raspberry Pi to a huge server it has a very small footprint only 20 megabytes and it has no dependencies it's like just a C plus plus project and that's it is very feature Rich we speak a why we have a a very large sort of SQL dialect we have all the features that you would expect from a normal from a modern analytical SQL engine it has Integrations with python a lot of other languages Arrow integration and can for example directly read parquet files ductibly is fast we have a state-of-the-art vectorized so-called query processing engine and that comes directly from the sort of state-of-the-art research that we did at the research group in Amsterdam for example daktibi can also automatically paralyze queries so you give it a query and it will figure out how to use all your Hardware resources duct abuse free it's free as in free beer it's MIT licensed do whatever you want build a company on top of it it's really up to you but let me zoom into a bit into some things that make that to be different drdb is not a client server system and then I know that is also something where it's a bit of a departure from what people are used to all data systems Under the Sun are client server and this has been unquestioned since 1984. but Dr B is in process meaning the database runs directly in application and it means you can directly access data in the application and in fact I'm going to show you a quick demo on why this is amazing here I have a python script where I create a data frame with a billion rows it's about eight gigabytes in memory and now I can spin up a ductib instance in that same python shell because it's in process and I can run a query and it will only take 160 milliseconds or so to go through these billion rows and compute the average and that has to do with yes the data is the database has a state of the art engine but it also can directly access the data in your memory of the Python process but let's zoom out a little bit how does duct TB fit on a typical architecture we have our data Lake we have analysis cluster and we have the laptop with the analysts well very often the first step is that we do ETL we take the big unwieldy data set and we transform it into a bit more manageable data in like data Lake formats or parquet but meanwhile the analyst is idle and then we do the analysis project where we do explore and we try to find out something about the data and we use the cluster as well and this works really well but it creates contention with other jobs running on the cluster and it slows down analysis with duct TV you can actually move the exploration and analysis on the very laptop that the user sits in front of and it has one real Advantage is there's no contention you have your own laptop it reduces latency and just generally reduces load on the cluster everyone's happy let me show you another demo here I'm now taking the dacdb either the famous New York Taxi data set it's about it's about 90 gigabytes of data it's in parque format and it's about 1.7 billion rows across 50 columns and I'm just using dacdb in this shell here to you know set up The View with the trips and now I can run for example very very basic query where I group the number of very accounted number of trips per cap type and this query has to look at all this data in this in this huge data set and it finishes in under three seconds seconds on my two-year-old Macbook so that is really something that you wouldn't have thought that you can run these huge data sets on a normal MacBook right you can also have a much more complicated query for example in disk query we compute the number of trips per amount of passengers the year in which they happen at a distance so this is a bit more complicated and I'm not going to expect you to read the SQL query but even this more complicated query will finish within under 10 seconds but how does duct TB fit into the machine learning space well we are not doing machine learning we are database people we're simple-minded but as you probably know the process of preparing data for analysis is the one of the things that you spend most of your time on if you want to do any sort of ml problem and active actually fits really great into that process because I mentioned you can pull a parquet file you can do a lot of reshaping wrangling you know figuring out what features are required creating new features all that you can do all that in duckdb and then we have Integrations with systems like pytorch or tensorflow to directly ship the results of this reshaping you're doing to these libraries and the cool thing is that because duct TB runs in process we are already in the same environment that the ml library is also going to run in which means the transfer is near instantaneous so now I hope I have given you some reasons to check out duckdb um here has some pointers but with that I thank you for your attention [Applause] [Music] there must be something in the water because llms are the Talk of the Town if anyone has tried if anybody has tried building an llm-powered application they have very likely used flank chain it's incredible that this tool that we all depend on and use was only open source less than eight months ago it is one of the fastest growing open source projects out there and had more than 2 million downloads last month alone so please join me in welcoming co-founder of langchain Harrison Chase [Music] thank you Brooke for the introduction and thank you databricks for having me my name is Harrison Chase CEO and co-founder of linkchin and I'm really excited to be here so what is LinkedIn link chain is an open source developer framework for building all applications we started as an open source python package in October of 2022 right in between stable diffusion and chat GPT and then we quickly expanded to have a typescript package as well we kind of came about right at the time where there was a massive increase in the number of people interested in building applications with language models and so I think we've been really fortunate to have an amazing Community that's helped build with us and on top of us and so I think that's reflected and these numbers are a little bit out of date we have over a thousand contributors and so I'd also just like to take this opportunity to thank everyone who's contributed in part to Lang chain so yes yeah that's that that's the most important part to clap at so um what is link chain so I think of the value props of Lang chain in in two separate ways one we have a lot of components and these are all the building blocks that are necessary when building llm applications and so at the core of llm applications are the models themselves and so what langtune provides is a standard interface for interacting with over 40 different llm providers hosters everything in that in that Spectrum um we also provide tools for managing the inputs and outputs of these llms so at the very basic level the inputs and outputs are strings but of course A lot of people are having more kind of like structured inputs so you have a bunch of different variables and user inputs and chat history and few shot examples that go into this this what ends up being a pretty complex input to this llm and so we help manage all of that state and on the opposite end we have output parsers which help take the output of the language model which is a string and parse it into something that's useful Downstream because the main the main way that we see people using link chain is is not just to call a language model but to use it as part of this system and so that naturally involves connecting it with a lot of other components and we have a lot of Integrations with those other components as well so document loaders we have over 125 different places to load documents from and and then split them into chunks and this is very important for when you want to connect these General models to your data and enable them to answer questions about about your company your personal files we we have an integration with over 40 different Vector stores again all with a standard interface so that you can start with a local Vector store and then quickly transition to a cloud Vector store when the time when the time comes so that's one of the value adds of link chain which are these little building blocks basically which you can assemble in various ways to build an end-to-end application and then on the other side we have use case specific chains and so we see a bunch of common patterns in how people are building all on applications and we put these into chains to do question answering over documentation question answering over CSV or SQL or any type of tabular data as well adding a natural language layer on top of apis so you can chat with your data that are behind apis extractions so extracting structured information from from unstructured text and a variety of others and so to summarize link chain I think of has two two value props one are these individual components which you can easily use and assemble and and are in are interoptable to build your own applications and then these chains which are easy ways to get started for the rest of the talk I want to talk about three of the main areas that we're really excited about at link chain so the first one that I want to talk about is retrieval and so the problem that retrieval solves is that language models like GPT only know the data that they were trained over and so that's useful and they can answer a lot of questions but when it comes time to using them to answer questions about recent data or your personal data they they can't do that by themselves and so a popular technique for for allowing them to do that is a technique known as retrieval augmented generation where you first do a retrieval step and you provide that information in the prompt and ask the language model to answer based on on that data and so the benefits here are that you don't need to retrain a model and so you can use any of the the commonly available apis off the shelves and then this also helps ground the model and reduce hallucinations so it doesn't make things up so the common workflow that we see for retrieval is when a user question comes in we first do a similarity search and retrieve from a vector store various docs and various chunks of documents that are relevant for that question at hand and so behind all of this there's also a separate ingestion process and this is where the document loading comes in play where you can load document from let's say your notion you can then split it into various chunks and the reason you split it is so that you can then retrieve only the most relevant ones you then create embeddings and you put them in a vector store you then pass them into a prompt and ask the language model to answer based only on that question this is a this is the standard solution and it will get you about 80 percent of the way there what we've seen and what we're really interested in is advances on top of that advances that help go from that 80 percent to a more reliable 95 and some of the edge cases that pop up that necessitate these advances are things that happen when you have conflicting information or information that changes over time or when you have queries that aren't just about the semantic nature of a document but also refer to some aspect of the metadata and so we have a lot of different layers on top of the basic similarity search inside of link chain to help get started with that another area that we're really interested in is agents and so the agents is a bit of an overloaded word everyone kind of has a slightly different definition the one that I like comparing agents to chains is that chains are a sequence of predetermined steps that you have coded in in code while agents use the language model as a reasoning engine to determine what actions to take and then it goes in and does that action offloading any any knowledge or computation to that action it gets back in observation and then it passes it back in and so the the standard solution for an agent is effectively a while loop where you ask the language model to think about what to do you then go execute that action you get back in observation and you repeat that until the language model recognizes that it's arrived at its final answer that it's completed is subjective and so so we have some infrastructure around that and then we also have various prompting techniques like react which is a paper that came out last October stands for synergize in reasoning and acting and it's it's designed to more effectively enable language models to think about what to do and take actions and so we have these as part of the standard solution and standard offering and Lang chain we're also really interested in advances in agents agents are one of the most fast moving spaces and so we're paying really close attention I'm really excited by plan and execute style agents where instead of going one step at a time you do a planning step you then execute the first step in that plan and then you return to the planning step and and kind of adjust the plan from there and this helps with longer running tasks and so this was this was heavily inspired by a lot of work in the open source by by projects like baby AGI actually and things like that another area that we're paying really close attention to is tool usage so tool usage involves the language model taking an action and so this started off where tools were functions that accepted a single input and that input was just string and that was because language models back in October were only really good enough to input that single string now they're getting good enough where they can input more complex structures with multiple function calls and so we're working to support that and push that forward the last area that we're really excited about is evaluation and we're excited about this because right now I think there are a lot of people prototyping and there's not enough people putting things in production and I think the gap between prototyping and production exists because it's really hard to get the more complex applications reliable enough to a point where you can trust them in production and part of that is evaluation and gaining that confidence and so there's two main issues that we see with evaluation for LM applications one is a lack of data and then one is the lack of evaluation metrics you don't start with data with a lot of these LM applications you start with an idea and these language models are fantastic zero shot Learners and you can get started really quickly but then how do you know how well it's doing and then on the evaluation metric side traditional ml metrics like accuracy and MSE don't quite cut it here and so one of the things that we're working on and we're excited to announce in a few weeks is a platform to help with this help with enabling both the collection of data as well as the evaluation of it and evaluation of it comes in a few different forms there's there's traditional metrics but it's also just making it easier to visualize and understand what's going on the most common way to evaluate is the vibe check which which sounds bad but it's actually it makes sense that that's the best way because it's really complex and allowing people to gain that insight as to what's going on is important and so we want to help with that as well as as well as more advanced metrics we also have a ton of Integrations with databricks so we support the the LOM endpoint we support ml flow and we have an integration with spark both chains and agents where you can interact with your data and the best part of all of this is that you can run link chain from inside data bricks and so that leaves me with the question we're here this is this is a generational moment to build things with ML and Ai and so I hope that that you guys take advantage of Lane chain take advantage of databricks and and what are the amazing applications that you guys are going to build I'm excited to see and and please share them so that's all I've got thank you guys for having me have a great day [Applause] [Music] thank you thank you Harrison Lang chain loves databricks and databricks loves Lane chain the databricks assistant which is live check it out was built using light chain in order to power these llm applications though we need a framework to build our language models and according to papers with code the most popular framework to do that with today Pi torch for our next speaker we have the former head of Pi torch who's going to talk about the past present and future of AI for Developers so please join me in giving a warm welcome to Lin Chao CEO of fireworks [Music] [Applause] [Music] hi everyone I'm Lynn today I will talk about developer Centric AI the past present and bright future I have always been passionate about building awesome tools giving to developers to do the best in their jobs when I started my career I built one of earliest in memory coloring databases fast document store with inverted indexes and many data tools by logging metrics experiment platform and so on and so forth I moved to AI about six years ago seeing a huge surge of data volume driven by AR workload a LED pytouch team at meta accelerate AI researchers in open source and bring huge production workload in production at metascale after having done all this I started fireworks on AI with link the mutual and awesome group of Engineers continue our journey to bring awesome tools to broader set of developers here including all of you in this audience and help you to be the most equipped proliferative creative in this new AI generation fireworks aim to accelerate product innovation building on top of dnai we offer experiment and production platform where you can get access to the most data art research of Open Source models bring your data and use our previewed tuning evaluation recipes customize your model deploy easily into our inference Services optimize for cost to serve and then we free you up to focus on your product Innovation ideation and iteration when your product idea is ready to deploy full speed full-scale to production you continue to stay with our inference service and go from there before I talk about more details of products and why us let me start with where AI Journey began Ai and deep learning is our new not new concept more than 30 years ago yellowcon as one of the pioneers of deep learning published a paper in 1989 back popular propagation applied towards handwritten zip code recognition the model still reads very modern today so on that point for many years AI remained as a research topic researchers need a maximum amount of flexibility to innovate and maximize model quality most efficiently towards the Trinity that they have when we started pytorch was started for researchers to export and invade faster and we observed around every three years major model architecture disruption happens and every few months significant incremental Improvement on model quality happens the pace of value created by research has a Slowdown but accelerated in recent years what is pytorch you can think about it as numpy for AI accelerated by gpus tpus your accelerators of course it runs on CPU as well pytorch use model as code philosophy for dynamic neural networks including Dynamic shapes control flows those are essential for air researchers to disrupt building disrupting building blocks in AI why model as code is so good because programming first program first structure is the most open and composable as opposed to rigid closed systems with hidden functionalities many research institutions including open AI switch over to use pytorch a few years ago for its flexibility to accelerate their Innovation speed so many researchers become entrepreneurs or move on to deploy models in production production deployment shift the focus on system performance stability versioning scalability cost to serve and so on as you can see there are quite natural tension across these two cohorts of air Developers researchers naturally gravitate towards breakthroughs and break current Paradigm to new ones and production developers optimize for current paradigm pytorch continue to speed up researched production transition by ease the tension in between and that's my focus in the past six years one of the most important use cases we innovate on is ranking recommendation because it's everywhere in our Digital Life python release recommendation libraries last year driving more than 10 times performance Improvement and memory utilization reduction it makes the large complex reconditioned model practical to be deployed in production we continue to to bring more awesome models on mobile devices which is more resource and power constraint it requires order magnitude Improvement of reduction of python time size and latency Improvement as you can see there are many cool features we enabled through this on mobile products pytosh is today deployed very broadly in production driving more than 50 trillion influence per day on servers in more than 70 billion in first per day on mobile phones pytorch application industry has penetrated across many domains from content standing of language vision speech to a complex task of ranking recommendation robotics to a high Precision task of medical cancer diagnosis treatment drug discovery many variations of autonomous driving vehicles in package build a Vibrance Community as we can see from the interpretation of archived people covering both research and production a constantly growing over the past years towards a dominant position as Janai is the most interesting topic in this Summit most of gni models are written in pytorch so we are actually going through a very interesting model development paradigm shift now from before where a lot of individual developers or institutions are spending lots of Energies collecting curating cleanup data and training a particular model architecture from scratch and then deploy into production to now you can easily customize your model using much less data as you can see from this Paradigm from bottom up there are many Foundation models already pre-trained and the race released based on public data available on internet and then you can specialize those Foundation models with your domain specific knowledge for example for finance legal health care and so on with their custom with their company proprietary data then you can further customize towards your company specific product by building your mode and the further along you can even provide a personalized model for example building avatars or air assistance to make your product even more interesting so we have a foundational shift right now because of Jenny I in AI development you still get the state of our models from research but with much less data you can start to customize and build interesting products and do product exploration much much faster one of out of many product ideas you tried getting to full-scale production and your Production Drive more data pushes better model quality improvement and you experiment and explore more product ideas in this quick iteration cycle is new as you can see in AI developer cohort we have a growing a new growing product developer coming up like many of you sitting in this in this hall right now so fireworks is created to accelerate product Innovation building on top of jnai by solving many of the challenges we heard from many of you first of all the challenges is there's a huge amount of dynamism of General models and there's a start competition of many model providers constantly improving model qualities one after another it is very concerning to many people how do how do I manage this ever-changing air landscape and which model should I use and stick to as you can see from the right hand side of the diagram with increasing timeline the intensity velocity of model Innovation accelerates second business task is never simple it usually is a complex business workflow that needs to be decomposed into one or multiple air models and which layer in this hierarchy do you interject and there are many interesting business problems people coming to asking us how to solve that for example I want to use natural language as the interface to retrieve precise information from my company internal documents or precise information from my product catalog or I want to automate my construction workflow to minimize timeline and minimize cost or I want to create diagram charts at Scale based on my data this is the first mile problem and when people get slowed down not knowing where to start more specifically within each model family usually there are different model sizes being provided from small medium large sometimes to extra large and you can choose to zero shot access the base model or prompt tuning to get better quality or fine-tuning user data to get even better quality for specific tasks you can actually fine-tune a equivalent quality or better model with a smaller size so which one would you choose of course AI January AI is expensive because it's computational and memory intensive in the past two to three years AI research has been pushing for bigger and bigger model size until chinchilla law get established that there's optimal token to parameter ratio about 20 to 1 but still most people are using around 7 billion to 13 billion model parameters and that's much more bigger than traditional machine learning algorithm for example of XG boost many companies can slow down or stuck because of the um the intensity of the computational cost and quickly run out of pre-planned Hardware budget so we think the competition in the JNL model space is actually great because the party developers are machine learning Engineers can write a free wave of ever Rising model quality there will not be one model provider dominating the whole Space there will be many of them keep advancing the quality altogether fireworks want to enable you to continuously testing and evaluating the best model to incorp into a product yes continuously because the Cadence of change here is extremely fast second cost so we love this because we have been driving customer service performance in general performance optimization for python for many years at hyperscalers we hyper optimize the cost for fine-tuned models fireworks today offer at equivalent model size five times cheaper cost compared with open Airbase model and 30 times cheaper cost compared with their fine-tuned models that's a whooping cost reduction we also quickly optimize latency for open source models for example a Nissan one that we deliver more than 10 times latency reduction based on the original implementation this requires deep hydrogen time optimization and integration with the best kernel as well as carefully crafted scheduling orchestration design and optimization all coming together we also understand to accelerate your product exploration you need to do a lot of concurrent parallel experiments at the same time and oftentimes with increasing number of experiments it increases your memory footprint as well that means you have to pay more we want to take away that concern for you as you can see we can effectively control the memory footprint with increasing degree of experiments that means with fixed budget you can run a lot of more experiments at same time in your time to Market will significantly be reduced tying all this together fireworks provide a rapid experiment in the production platform we give you free access to the state of art research with your own data you can use our pre-built fine-tuning and evaluation recipe you can change it adjust adjust the W As You Wish run the fine tuning on-prem or on cloud as you wish and their custom model will be deployed to our general AI inference service with minimum cost to serve you can then focus on your product iteration and exploration when you like your idea and want to grow into full production scale do that within the inference service without changing to anything else today we build our inference service on top of gcp AWS and quarries we plan to expand our cloud provider partnership quickly down the road our API is fully compatible with open air API and lynching so if you you don't need to change your code much to use our services if you feel the challenges I just mentioned resonate with you and also want to know more about fireworks find me after talk and also find an awesome fireworks AI team we can give you a demo and give you more information about worldwide building also stay tuned for many more exciting announcements coming up soon thank you looking forward to seeing all of you [Applause] [Music] our next speaker is a Pioneer in the field of computer vision he was formerly the head of AI research at Facebook in Menlo Park and as a distinguished professor at UC Berkeley in his talk he's going to talk about the latest developments of modern AI applied to robotics please join me in welcoming jitendra Malik to the stage [Applause] hello so I want to talk to you about the sensory Motor Road to artificial intelligence but first we should know something about natural intelligence and natural intelligence to make sense of it we should think of it in the light of evolution so something like 540 million years ago we had the first multicellular animals that could move about and moving was great because that means that you could get to food in different places but to get to food you needed to know where the food is which means meant you needed Vision or some form of perception and these two abilities go together and there's this great line from Gibson we see in order to move and we move in order to see and so if you think of the brain or the nervous system of a fly or a fish or a frog it's basically this connection between perception which could be Vision it could be hearing to action and action is moving about in this case so if you zoom ahead to like five million years ago which is the when the first hominid split off from primates you have this additional accomplishment which is that we start to walk on two feet which means that the hands are free to build tools ah make tools uh and and then you get the Advent of dexterous manipulation and planning and all the rest of it and then the last big development is of course for modern humans like hundred thousand years ago or something which is when we have language which is uniquely human and abstract thinking and symbolic Behavior but what's important to keep in mind is that most of the brain is devoted to perception and action and connecting the two and if you think of the entire evolutionary history as being 24 hours of intelligence languages in the last two or three minutes that's all it's very important but it's only the last two and three minutes and we in this audience I don't need to say much about language models but it's they're incredible and they can do the I'll pick one line they can pass the bar exam at 98th percentile incredible so given such amazing accomplishments what I want to connect you to is the fact that we have so much trouble on another side of AI which is the robotic side of AI or self-driving cars I have been in this field for 40 years we have had self-driving cars for 30 years cars which drove across Europe from Berlin to Paris and across America and so on and there's been a lot of hype about self-driving cars and we'll get them we'll have the cars but think of how hard it has been and it's something which so we can pass the the law exam which takes years and something that a high school kid of 16 wait after 20 hours of training is good enough at we are having trouble with this I can make this problem even harder so think about what a 12 year old kid can do right a 12 year old kid in a kitchen with knives and forks and ladles and so forth can do all these kinds of tasks and no robot today can do all of these tasks okay this is incredible something easy we can do and this is something that people in AI have known for a long time it's a paradox right a law exam the bar exam which takes years of study is hard and cooking an omelette it's supposed to be easy but actually it's the other way around marwick had this great line which is that things like chess are easy language is easy what's hard is what a one-year-old can do and Steve Pinker has a beautiful line for this which is that the hard problem what we have learned from years of AI researchers that the heart problems are easy and that the easy problems are hard and as the and then he goes on to say that the gardeners receptionists and cooks are secure in their jobs for the years to come and why so the question is why and model X argument was to do with reverse engineering and it was in the area of Designing Ai and there the question was okay what we have what has emerged through hundreds of millions of years of evolution is much harder to sort of reverse engineer I think actually the argument is slightly different we know how AI has been achieved it's largely been through uh deep learning applied to huge amounts of data and the kind of data that we have what's the kind of data we have we have huge amounts of data for language models why because everything is on the web right all the books are on the web Wikipedia is on the web read it for on the web GitHub is on the web this enables you to train these these models so all this data is available explicitly think of what's the data needed for sensory motor training you need to know what images I take and what are all my muscle commands and what are my neural activations hey that's pretty personal I'm not uploading that on the web okay we are not going to get all that data in huge amounts on the web we could get parts of it we might see what the images are but we're not going to get all of this and therefore we will need new clever ways of solving this problem it will need AI it will need learning and I'm going to take you through a little bit of how we can do this and I work a lot in robotics and this is one of the robots that we trained in my group and it's solving problems its footwears stuck against The Rock and it managed to go through uh this robot by the way is blind it has no vision it's okay and you see it walking downstairs it's not even aware of the stairs but it manages to stabilize here's another example which is on lose mud pile and at a construction site and so on and so forth so you need a lot of Versatility so these problems are actually hard and let me think of it more formally uh computer vision is like pattern recognition and we the basic challenges generalization so we can't have a formal definition of a chair but we can train a classifier for chairs by giving lots of examples when we come to these motor control tasks it's a different game one part of it is handled by classical control theory which is robustness to disturbances you do feedback control there's the second part which is which is actually the more interesting or harder part which is that adaptation to these different conditions so I showed you this setup where this robot dog has to walk in all these different terrains so adaptation that's very important so we can use classical control theory techniques to Train control controllers for particular situations and for example Boston Dynamics has a lot of work demonstrating these kinds of controllers but where Ai and machine learning can come in is can we build one policy to walk in all of these situations one policy which figures out automatically which situation you are in and then it works in them and this is a work from my group called rapid motor adaptation where we figured out essentially how to do this and I'll just take you through the big idea so the big idea is we train in this robot in simulation and this base policy is like figuring out how to change all the joint angles and things like that and there is this variable Z which you can see in that diagram and this Z captures some aspects of the terrain as some low dimensional vector and if we knew the Z we have this policy which will do different things so it does different things in sand versus and hard ground and so on and so forth okay but we how do we do that in in the real world we need a way to estimate that itself and for that we sort of need to go meta so we have an adaptation module which looking at the past history figures out how to uh what does he must be so the intuition is something like this that when I walk if I walk on hard ground I perform certain actions and then there are consequences of those if I do the exactly the same thing when I'm on a beach it's going to be different because when I put my foot down and I try to lift it up it's not going to lift up so easily and that kind of signal from the state and actions over the past one second or half a second I have the signal and uh and and that's basically it I mean there are some details obviously but I'll show you an example here so here is an experiment which uh my student Ashish did where he's pouring olive oil and on this waste of good olive oil on a mattress and then he's going to take the robot and if you look at the legs of the robot he's got like plastic socks to make it hard and then he's going to make this robot walk and what happens is that it it starts to slip right let's do it in slo-mo so what's happening is that it has some estimate of the extrinsics this Z vector and that estimate is wrong because it's slippery now what should happen is that over time when it walks it estimates this and that's being done by this adaptation module and once that estimate comes through it works out and then it's recovered I'll give you another example now it's a much harder problem I've got this robot which has got a vision system in it and it's going to walk in much more treacherous Terrain so now there's a camera and it's everything is on board it has no Advanced knowledge of the terrain and it's using similar techniques notice that this robot is a very short robot compared to the heights of the stairs and it's yet able to manage in these conditions and it has no prior knowledge thank you and this is slippery ground as well as a slope so and these ideas apply to other applications besides walking thank you thank you and there's one which is dexterous manipulation so if I want to cook in a kitchen I need to be able to manipulate with my multi-fingered hands and here's an example of that similar idea estimate what the situation is the so in this case there are objects of different size different shape different weights and and I mean that's what this says and these are being estimated online and uh so for example the shuttlecock is very light it's only five grams okay and some of the objects are heavy okay okay there's this empty bottle a Rubik's Cube very importantly it's exactly the same policy the robot is blind it has no prior knowledge of what it is trying to control but only from the appropriation what it feels in the fingers it is able to do the right thing and I have more examples here so so I think that this is the future I think we have to machine learning and AI is essential for the success of Robotics because we need flexibility and we need adaptation but I want to conclude in the last two minutes that I have with some general philosophical remarks so I entered the field of AI 40 years ago for me the success in the last five years is incredible I would never have thought we would have get got so far in that what we did in the last five years where we were five years ago versus now and broadly speaking deep learning in the last 10 years but how do we do the rest I think robotics is very important sensory motor control is very important without that we have not achieved intelligence there are these ideas for this which go back to Alan Turing Alan Turing who's like the father of computer science in a way and he has this paper from 1950 which has the Turing test but it has this great line instead of trying to produce a program to simulate the adult mind why not rather try to produce one which simulates the child's so it's essentially a program of learning but learning with stages the way children learn and we not know a lot from our psychologists colleagues about how children learn children go through various stages of learning there's this multimodal stage this kid in the Crypt she's playing around she's poking at objects so she here sees it she hears the sound she puts things in her mouth her motor system is being activated all of that is training data all of that is being used the kid does experiments so Allison gopnik has this book called The Scientist in the crib so when this kid when your toddler is being difficult and throwing food down you should say she's actually a scientist in the crib she's conducting an experiment from which is building models of the world around us and and this is very important and then finally this child at the age of two you take them to the zoo you give them one example of a zebra it works and our programs we need to give them thousands of examples of zebras and uh and then at the age of 16 we give them 20 hours of training and then they can they can drive right so we I personally believe that this developmental story is going to be needed for all of AI and the psychologists have actually told us what these steps are multimodal incremental physical explore be social learn from others and finally use language so language is very important but it is in a way the crowning achievement of intelligence and it should be built on the substrate of physicality thank you very much for your attention [Applause] please welcome to the stage co-founder and Senior vice president of field engineering arcelon tavakoli [Music] good morning everybody it is my distinct pleasure to help Host this fireside chat today we've clearly been spending a lot of time talking about AI you know where is it going what's the future how do we get here what's next and my guest this morning is somebody who has been living in that world for a very long time and needs very little introduction please give a warm Round of Applause to welcome the former CEO of Google Eric Schmidt foreign [Applause] [Music] hi everybody good to see you thanks for coming I'll be honest I've been at databricks for 10 years now and uh my parents were the most excited been that this morning they're like they didn't care about anything else you'll be on stage with Eric Schmidt I was like yes well you know I was here before I was here 30 years ago on this stage and I introduced Java right here yeah so quite an accomplishment do do the math so there's obviously a lot of directions uh that we can go into I'm pretty sure I'm gonna get yelled at to get off stage before we can get through everything but you know one of the topics that everybody is interested in is just the speed and fervor which things have come on the generative AI side and you've been living in that world for a while now so curious to just hear your perspective why now What's led to this kind of interest in Drive well what's the value of inventing a new kind of intelligence foreign probably pretty high what's the value of improving every single business process Communications process entertainment process educational process in the world pretty high how big I did a report for the US government which said it was a 40 trillion dollar business that's big enough to get everyone to play and the other thing that's interesting is our industry when I started which was almost 50 years ago we were not very good at talking about ourselves we were actual nerds we were not actually nerds with shine if you will and the industry has changed and we've gotten really capable of hyping ourselves to the max right and I can give you the data of adoption look at the rate at which chat GPT was adopted versus Gmail as a simple metric but the fact of the matter is it's both better hype better excitement and compression of time one way to understand this is that before the internet existed your company you actually had to sell you had to call on doors and send CDs and so forth and so on and the internet created at least for digital Goods an immediate instance success history which is now what's driving this there's this sense that you can go completely non-linear with something successful the other problem is that the compression of time means you have almost no time to get it right and small differences in timing can have huge statistical outcomes based on when you started and when you got there which is why everyone is in such a rush to get funded to get people and so forth everybody kind of understands do it right now okay so there's a there's a little bit of that now it's very interesting somehow you know generative AI all of a sudden has become all about the model everybody talks about the model and um this big question we get asked a ton about it from customers it's been a topic where I think it's fair to say for many generative AI became synonymous when chat GPT and open AI came out and so uh there's this debate is the world going towards this part where there'll just be a small handful of these large foundational models or is it going to be more towards whether it's open or specialized model is it a balance we get asked a ton obviously uh you know I've seen you talk about it how do you think about what the future is the need of the different kind of types of models and how they'll be leveraged so about a month ago a number of us wrote a blog post which tried to answer this question yeah and the answer of course in Tech is both you're both going to see these incredibly powerful very very flexible what we call Frontier models and they're going to get used in business processes but you're also going to see incredibly powerful specialized models both will happen for different reasons if you think about it as a business so one of you is trying to build a company that uses advertising but you have an idea to generate the ads for the customers instead of making the customers generate the ads that's a good idea and by the way Google has just now announced they're going to do that good job so you do that do you really want a fully functional learning changing and so forth powerful general purpose model or will you be better off served with a specialized model that does a better and better job and I think in that case you would want the latter if the questions are more open-ended you're probably going to want the more General model since we haven't proven the business models of either these right we literally don't know exactly how they're going to get monetized yet which by the way is a testament to the world's Financial system that we could raise billions of dollars without having a product plan a revenue plan a product price and an identified customer thank God for the financial system and right we're all here because of that so so I think you're going to see both um and one of the most interesting questions is the rate of diffusion from the frontier models to the open source models so if you take a look gpt3 is now the functionality is largely available in the equivalent of alpaca um that's roughly two to three years in terms of diffusion from a very expensive very powerful model to something which is generally available and effectively free yeah and I think the question becomes as we keep getting more and more powerful at what point is it good enough like for for the different uses well the correct answer is it's never good enough in software yeah right we still we have the M1 the M2 the M3 the M4 we have five nanometers five three nanometers two nanometers it's never good enough but somehow we find a way to use up all of the the software it's the old rule you all probably don't remember this it was a Grove giveth and Gates taketh away right was the saying that Intel would add CPU yeah and Microsoft would immediately use all the all the hardware for the software that they were building on top this is you know 30 years ago it's still true okay um so we were talking about this you know briefly backstage there's a lot about the models we spent talking about the specialized models and all that um in this world how do you think about the importance about things such as like data and governance and data quality in terms of driving improvements and Effectiveness in the use cases around it yeah what's interesting is if you look across the fields that are using AI people people talk about the algorithms so if you think about it you need what do you need you need Hardware you need scientists you need programmers who can build scalable systems which is not the same thing as scientists and you need lots and lots of data in many fields the most interesting data is data that is being synthesized right so for example if you have a physics model and you can act you know the physics is the physics you know physicists are always right in in my in my world they can essentially simulate and therefore generate the data that's needed for training there are plenty of other examples where people will do essentially sampling of the data so they'll for example sample a series of Q of questions and answers and then they will perturb and generalize those for training all of these tell you that data is the key and it's important that the data be curated you I wish that you could just sort of run these things over every piece of data regardless of structure and so forth and I think this is sort of why you guys founded databricks as you were trying to do this right and you've done it super well so you're one of the components you need to also be helping people hire the right technical people get the hardware and so forth to make it complete got it and so in the role of that you you basically just touched upon it of data right now the other aspect you talk about how important will human feedback be in looping that in in order to basically make sure the models are improving and AI is successful ultimately overall well rlhf um is a really cool idea when I first heard it I figured it wouldn't work but it turns that it works really really well and there are now newer new techniques which allow our lhf plus Laura basically quick adaptations of the model which allow you to fine tune using relatively standardized open source packages that are available on GitHub and so forth so that has led to this enormous explosion of variance and so the most likely scenario in my opinion we'll see is you're going to get a pre-trained model that's quite good a base model and then you're going to see every conceivable combination in the pipeline right rlhf various other things synthetic data synthetic training evaluation and so forth to get there the next thing that happens after that is the ability for these models to call to some reference point some ground truth source and there are startups that are working on calling when it gets confused which they always do calling out of the model into something to get a reference point and then put it back okay um maybe maybe Switching gears a little bit slightly from that we've been talking a lot about the potential the opportunities for AI uh there's definitely a lot around there I don't know if fear is the right word but uncertainty of where it's going and that kind of coincides with discussions about regulation and what's regulation going to be how things are going um how do you see you know do you see us broadly getting alignment on how to regulate Ai and the path forward or at least the challenges that are there categorizing those and what we should be how we should be thinking about those going forward well it's interesting that every single politician that I speak with every single leader I talk to is now an expert in Ai and they know nothing about what you all are doing so maybe that's always been true but it's sort of alarming to me and so the first thing I always tell them is um do you remember that movie where the robot gets out and then the female scientist slays the robot at the end of the movie that's a movie okay that's not we're not building Killer Robots yet right and that usually sort of disarms them so when you talk within the technical people within the industry who are I think the only people who really understand what's going on there's a consensus around three reasonably clear short-term threats that are important the first one is the one around misinformation and disinformation and we can talk a lot everyone understands what that is and everyone understands it for the more generative AI is going to be used enormously for that kind of stuff so we've got to talk about that the second thing is the ability for these systems to do various forms of Cyber attack and the ability for these systems to do various forms of Bio basically bio threats of one kind or another and the consensus is that today so first place AI alignment is a term where how do we get these AI systems to follow human values AI safety means what are the guard rails that we put on these systems if you look for example a gpt4 open AI spent about six months they had a whole team who basically using rlhf and other techniques constrained the model yeah and those constraints apply both at the API level as well as at their app level so people kind of understand that model and the and there's always this question of how do you how do you test for safety issues that you don't know yet without deploying them so there's a general fear that these systems when they're when they're launched will not be fully tested because they'll have some emergent behavior that we have not yet seen and therefore we can't test them so that's problem number one problem number two is that because of synthetic biology and so forth it's likely that these systems can accelerate at scale evil people and so if you sit down with Google long enough and you understand enough about biology you can probably get to a bad pathogen these systems make that somewhat more likely so we have to sort of think about that so the governments around the world all have variants of approaches to this but the simplest way to frame it is there are scenarios of extreme risk and these systems are going to get regulated around extreme risk I'm not talking about the things we always complain about about you know you know Johnny Johnny's dog ate you know Susie's homework kind of thing I'm talking about thousands and thousands of people harmed and killed from something yeah so uh you know maybe one or two more questions for me because I know we'll run out of time but it always happens we've I don't I don't want to call them kind of uh hype Cycles but whenever we get these technology there it seems like we go through transformational technology there's some fear there's some people who fear it there's some people who believe that tomorrow all of a sudden we don't need programmers anymore because generative AI is going to take up neither the two actually come to fruition um what's your sense of the timeline for when we will start seeing very very meaningful we're already seeing some of it but like kind of almost crossing the cast of meaningful adoption and leveraging of generative AI like kind of across all Enterprises and spaces so so the first use of generative AI is already with us which is programming and the first and most obvious use is in enhancing the power programmers and this shouldn't surprise you because every generation that I've been through that technology was first used by its inventors to make them more productive by the way I'm old enough this is what email was invited invented for right this is what we used Unix for way back when we used it to make ourselves more productive and I used to tell people inside the various companies that I ran that why don't you start by making yourselves happy which is really hard and then why don't you make your friends happy and then come back to me right that sort of works we sort of forgot that we decided that we could build arbitrary consumer products without actually using them ourselves and that's a mistake so I think the first use is programming the gains in programming are profound and it looks to me like half of programming can be sort of think of it as a doubling of productivity right and that's going to that's going to continue there's a whole bunch of startups that have even more sophisticated ideas around that I think that's the first one I think the second one is going to be in doing things which are in the normal course of businesses I earlier mentioned this notion about advertising why again why do I have to decide and argue through my tweet why can't the computer ask me what I want to talk about and generate a tweet that is guaranteed by its metrics to be the most viral right so if if I'm trying as a marketeer to have the biggest impact surely the computer working with me can help me make it happen so I think that all all of these systems where the compute the human is doing something that is against some goal that's administered by computers is probably the next one makes perfect sense to integrate that makes sense right we know what virality is because this is what we do all day whereas we do it 24 hours a day and you're sleeping yeah so we'll tell you how to do it and and again that's an improvement in efficiency quality and then the next things are these more specialized markets we don't yet know what the value of intelligence is we don't know how to price it but it's obviously High okay I like I like that set of categorization and we see something similar in terms of the use cases people are adopting now one of the big barriers right now is people think about you know llms and building them and using them has just been costs cost is a huge barrier it's what you know the cost to train them access the gpus no secrets why you know Nvidia is now a trillion dollar company uh you know I'm just curious how do you think that barrier of cost both for training for inference for Access for infrastructure you know and access to the models coming down over time to make them more accessible it's it's interesting that the that the training cost is extremely high and going up by orders of magnitude there are people who believe that the frontier models which cost on the order of 100 million dollars to train plus or minus will go to a billion dollars to train one of these things so that is a massive change from a software perspective in my career we've never seen that kind of price increase to do software which is what they are on the other hand inference that is the the ability to actually answer the question looks like it's going to become trivially expensive in other words incredibly inexpensive and so you end up in a situation where you train one of these models and your biggest problem is making sure it doesn't get stolen right you spent a billion dollars for these weights right which can basically be put on the equivalent of a hard drive you need to make sure they don't get stolen and used by your competitor or or whatever and so my guess is that it right now you have Nvidia leading it's it's a sore point for me because at Google we have this whole TPU strategy but 95 of this of the training is being done on Nvidia a100s and now h100s and Nvidia has recently talked about a successor next year uh which they cleverly call h100 next which sure looks like it's another acceleration in power yeah so what Nvidia did particularly well is they worked on what is called the fp8 cluster where basically it's 8-bit floating Point multipliers very quickly and they put a special module in that that helped them and now they in the h100 they have an integrated memory architecture that's stronger one way to understand that is that the the way this training works is you have this very large amount of data that's larger than you could put on on a computer it's essentially in the network on the in the data center and the algorithm is going back and forth and back and forth and back and forth for hundreds of millions of dollars yeah anything you can do to decrease Network latency bring the memory closer is a huge speed up we're still in the phase where these computer architectures are evolving faster than I've ever seen much faster than Moore's Law much faster than I saw CPU architectures because of this unique nature of these llm architectures and that's going to continue for a couple more Generations okay um we're about out of time one last question you know you've got tons of Enterprises uh sitting in the audience here if there was one peaky key piece of advice you could give them of this is what you need to think about in order to be successful and successful in harnessing the power of gender of AI what would that be well the small companies are run by people who are here because you understand that this is core to your business so a simple rule for a small business is you're not going to be successful unless you use AI in your model and a simple a simple way of thinking of your business is you have an Android or an iPhone by the way Android is more popular than the iPhone just for data and you have a network and then you have a fast you have a server typically in the cloud somewhere which is using AI in your business yeah so think of it that way you understand all of that for larger corporations and I talk to lots of the leaders of those what I tell them is the following you don't know what you're doing yet in this space I say this in a nicer way you don't know what you're doing in this space so and your team may not either so give them the following assignment for every business unit you have come back and show me a prioritized list of what you can do with generative AI all right just show me the list right and get them all together and most companies that have done this will come up with 400 or 500 ideas some of which are customer service some of which are Revenue enhancing some of which are security right and they'll come back and then the CEO says holy crap right look at all this stuff that I should be doing and then they realize they have no people to build it and then they have a crisis awesome well Eric I'm pretty sure I speak for everybody here it says thank you so much for making the time for this it was amazing congratulations to you guys and congratulations for everybody thank you thank you our salons again appreciate it sure thank you okay so one last announcement for folks uh this wraps up you know basically today's day of Keynotes a bunch of interesting really interesting breakouts save the date data and AI Summit basically 2024 will be back here at Moscone June 10th to 13th as far as I know we're the only conference scheduled that week if that changes don't look at me it's not on us right take care everybody cheers

Channel: Databricks

Views: 6,744

Keywords: Databricks

Id: un_telVvKiY

Length: 163min 39sec (9819 seconds)

Published: Thu Jun 29 2023