Artificial Intelligence: Boon Or Bane? | We The People

hello and welcome you're watching we the people I'm Sarah Jacob this week a letter co-signed by Elon Musk and thousands of others demanding a pause in artificial intelligence research has created A5 stop on the 22nd of March more than 1800 signatories including musk the cognitive scientist Gary Marcus the Apple co-founder Steve Wozniak they all called for a six-month pause on the development of artificial intelligence these signatories to this open letter key figures in the world of artificial intelligence they want now they say they want a training of powerful AI systems to be suspended amidst fears of a threat to humanity in this open letter their warning of potential risks they say the race to develop AI systems is out of control the letter asks AI labs to immediately pause for at least six months the training of AI systems that are more powerful than gpt4 so what's GPT 4 well developed by open AI That's a company co-founded by mask and now backed by Microsoft gpt4 has developed the ability to hold human-like conversation it can do everything from composed songs poems and summarize lengthy documents such AI systems which have what is called as a human competitive intelligence pause profound risks to humanity the letter claim now Advanced AIS they say need to be developed with care that's what the letter says but instead what we're seeing in recent months we've seen AI Labs locked in and out of control race to deploy develop and deploy ever more powerful digital Minds that no one not even their creators can understand predict or control so you have Elon Musk and top AI researchers calling for a pause on giant AI experiments in this open letter that says the current race dynamic in AI is dangerous it calls for the creation of independent Regulators to ensure future systems are safe to deploy so it all sounds very worrying what should we take away from this that's why we know you feel the same and you have these concerns and that's why we're doing the show we put together this show for you so who are the signatories to this letter well the signatures include the author you all know a harani the Apple co-founder we have Steve Wozniak we have the Skype co-founder John Tallinn the politician Andrew Yang a number of well-known AI researchers and CEOs including Stuart Russell Joshua benjino Gary Marcus stability aico Ahmad mosstick and researchers at the alphabets AI lab Deep Mind so the question we're asking is the world prepared for the coming AI storm joining us on We the People tonight we have Mateo Flora professor of Adjunct professor of corporate reputation International cyber security expert and AI expert and an entrepreneur joining us from Italy ankush sabarwal founder and CEO of core Rover Tech expert associated with the government of India for IRCTC Etc apps jibu Elias head of content and research at India AI That's the national AI portal a fellow for responsible Computing fellow for India at Mozilla and kazim risby founder director the dialogue expert on AI and a tech policy expert Mateo Flora let's start with you simply because Italy also made headlines this week because it banned chat GPT was banned in Italy over privacy concerns in this action by Italy's data Protection Agency is the first known instance of chat GPT being blocked by a government order can you just put this into perspective for us why what prompted this and what the reactions have been in Italy to this open letter by mask and many others considered to be AI experts well let's start from what happened the Italian data protection authority has imposed that temporary restriction on data processing for the Italian users of chat GPT as a result open AI is now subject to an official investigation regarding its compliance or not with the data purchase protection regulation as a result open AI has unilaterally blocked the access to Italian customers but this will in no way be sufficient to answer the question arised by The Authority that covers four main points first of all a data breach openai Express experienced data breach on March 20 exposing user conversation and information related to its paid subscript Subscription Service the DPA notice reveals concern regarding the absence of whatsoever disclosure about the collection of data the second one is uh that the dpas specifically highlights the lack of justification for collecting and storing massive amount of personal information that was used to train the platform itself so we are talking about training we're not talking right now in conversations done by user the third one is uh that the um the Privacy Authority found that the information given by Charter GPT did not always match the actual data of people and so there is an inaccurate processing and music representation of personal detail and the fourth one that's the easier one um while chat CPT is targeted to user above 13 years old as stated in the terms the terms of services however the DPA has raised concern regarding the absence of any kind of age verification mechanism which could expose Minds minor us do unsuitable information Etc the company has now 20 days to comply and to answer or face a penalty up to 20 million Euros or four percent of the global annual revenue whichever is higher all right uh all right I'm not sure how how one can actually enforce this and how one can live in as what we call the World is Flat but it's a great conversation starter what happened in Italy can we see this in other countries because that's one of the aspects of the letter raises we need to look at what kind of restrictions need to be put in place and are our governments on board but first of all I just want to understand get back to the bigger picture uh kazim risby can we just uh get from you um put this into perspective for us how has this open letter actually gone down um across the world uh in the US and the tech World in Silicon Valley because we know who signed the open letter right but interestingly or perhaps predictably Sam Ortman chief executive open AI is not among those who've signed the letter yet I think uh Sundar pichai satyana Della CEOs of alphabet and Microsoft they're also not among those who signed the letter so can we say this is the gospel truth this is seriously in issue and everyone needs to respond with lacrity to the contents of this open letter thank you so much I I was just trying to sort of respond also in terms of what the operator is talking and uh just a general conversation around chart GPT but also the use of AI globally and certain ideological questions that have to be answered I think what the uh Italian uh government is trying to do is understand and test uh charge GPT with respect to how it's complying with gdpr and with local data privacy laws I think that is very important uh and I think it also sort of uh brings us to the question in terms of what are the principles of Regulation we need to imbibe uh globally as well as in India yes so as to minimize arms and prevent risk of any technology uh bear in mind the fact that look it is an important technology and I don't think uh the right solution would be to say that look let's do away with it I think what we need is a lot of collective discussion and maybe the open letter is a start to a conversation in terms of how do we regulate AI going forward both on the level lot of corporates who are developing this tool but also on the part of the government right um there are three things which I sort of want to talk about number one is around bias and we understand that you know AI depends on data the more data you feed the more accurate it becomes but we also need to understand in terms of the quality and veracity of data which is being fed into the AI Technologies and how uh this data is being collected who are developing the technology I think we need a lot more diversity convolutions around diversity by Design if you see by Design when it comes to developing AI technology both at the level of development but also once the technology is being used in the public domain and our charging PPT learns from humans right it sort of uh takes the learnings and improves itself which is a great idea but at the same time you also need to be able to understand that how the technology is evolving our period let's say two years four years or five years and what are the upgrades that are coming on and how do we ensure that you are minimizing the harm to society thing catch up by the time you figure out what to put in place in terms of Regulation the technology is already moved on to the next version but ankush so we're talking about cons right like we know there needs to be regulation we know there are cons whether it's risk to a democracy because you have like what you call weaponized disinformation there's a loss of possible loss of jobs because of displacement of human skills education there's plagiarism there's devotee I mean there's cheating basically um this letter also there was a recent report from Goldman Sachs that said yes AI will increase productivity but millions of jobs could become automated so can we also just talk a little bit could I put you on the spot could you tell us what the pros are then for artificial intelligence uh thanks for asking this question to me you know it's like uh you know the analogy is uh say if you hire someone as coordinate to you and if that's spotting it rights to outperform you and he's generally intelligent or what could you do would you fire that guy uh right you will perform uh the best whatever that guy can uh but at the same time you build the safety net and goodness around that so in the latter I've read that letter if you see there are two aspects I think spoken are articulated very well I think the first they are very articulate saying hey the safety net then governance should be there yes absolutely the government should be there but uh but can we say hey so you stop uh the experiments around it is building a safety net and governance not part of the experiments not part of the cleaning well even if you do that how would you ensure that doesn't happen I think uh whatever is the genuine demand and after putting the safety net important that should be there but but I think stopping it uh I think that's that's that's not feasible and probably that's not going to happen uh but all right so you're saying don't throw the baby out with the bath water right let's I think at this point of time let's bring in now if you're talking about the the pros how it can be used for a greater good we have someone I'm very honored to have on the show Lena rafik who is uh all of 11 years old and she's sharing the platform with all of you man I'm glad this is not a manner because this 11 year old girl has introduced an AI application that can help detect uh eye disease by screening and analyzing your eyes so fantastic Leela Rafiq thank you for joining us quickly tell us about your Innovation and what it has to do with AI Leela you may need to unmute my opinion yeah should not be paused or feast because it's going to wrap rapidly and very quickly and it's very very useful for the human so you so you introduce an application it's called ogler iscan it detects four different types of eye diseases right and you can find out if you have a cataract and Arcus melanoma and the app comes up with 70 percent accurate results is that correct Yes actually yes so four diseases has four AI models currently and 70 accurate results all right just quickly tell us Lena how did you get into this AI when you were like you're only 11. were there concerns it was their family relatives Etc asking you know why do you spend so much time on the internet and you know how one has to be concerned just take us through what your what you face growing up well actually my family school is very supportive actually they gave me kgp uh Advanced knowledge gaining process for me to learn this at advanced artificial intelligence machine learning and coding yeah so they were actually very helpful and more supportive and this is what you want to do in in the future what do you want to do going forward well I want to uh I have goals and I have a I'm going to learn more and using that I'm going to fit solve many issues in this world all right you're gonna solve many issues in this world we need more uh young people like you with those kinds of aspirations who are able to make use of AI for what purpose perhaps it has been created to do more good in this world which Elias is that am I being naive we have new and capable regulatory authorities dedicated to AI that's what this letter says that we need regulatory authorities that are dedicated to AI we've had recently a number of proposals for the regulation of technology in the U.S in the UK in the EU where does India stand on this so I think uh you know India is India is in a very early stage when it comes to AI regulation but at the same time you know what Lena is doing right that's one good example of how AI is being used for uh the social good right for example in India we got only one eye doctor for uh um 200 000 people and now if you have tools like what Lena has developed I think a similar tool is employed by Tamil Nadu government so this is where AI comes in handy right like uh bringing down cost of Diagnostics or some of this areas in healthcare and agriculture it's making a lot of progress at the same time you know before I get into the regulation part I want to say a lot of this current uh situation is a lot of hype there even the open letter the people who wasn't signed like many of the top AI ethicists hasn't signed this open letter look at him that Libra who warned about this situation two three years ago to her paper stochastic parrots or a Yan Lake right so people are also pointing out the what happened in 14th century when typewriters came out the Catholic Church immediately put a ban off for six months right so that's one thing secondly uh we are nowhere close to calling it a digital mind you know we it's nowhere AI uh like human competitive intelligence not what we are saying is just language models and this uh model architecture called Transformers which is very good in predicting the sequence of characters or words which we find meaning in yes these these these kind of models can replace jobs I think CNET is using this kind of jobs to create articles now even I have played with a gppt2 even in India AI there are articles we purposefully got gpt2 to write on so those things are there but at the same time we have to understand that the level of we have to filter out the signal from the noise the level of hype out there uh now come to the regulation I think EU has one of the best models which are risk-based models right the more risk a particular segment and how for example Healthcare using AI Healthcare is more riskier so tighter regulations or or if it's look at some something um like a chatbot Channel they have a lesser regulation but now they are regularly as a result of this chart GPT and all so I think uh you know we are also on the right path we have the digital India build coming up we have the national program on AI coming up at the same time India has a different view India wants AI for all we want AI to empower our society like the bashani program we have which is a very good example or Angus has worked with uh many of these IRCTC and many many of providing these tools right the uh my Google help desk and so many programs the government has developed which is kind of bridging that Gap even a few days back Jensen Wong said this kind of AI will help India because now you don't need experts to make calls you know people like Lena or young people who don't have that kind of uh experience or who are coming into this space can you use models like charging to write codes you don't need to learn python or uh C or C plus plus to create program so in that way these new models will close the detail divide we have about the development and about how these things are turning out to be but I don't think anyone is gonna take this open letter seriously or gonna shut down the uh and how much will six months how much of a difference can six months actually make right uh uh Mateo Flora interestingly you have the Twitter Chief now Twitter Chief Elon Musk one of his many oftas one of the many hats that he wears he is among those who now want training of AI is above a certain capacity to be halted for six months he's uh he is the chief executive of Tesla so one of the co-founders of a company where basically uses AI for an autopilot system so that in a sense uh hypocritical in a way or can you should we look at it as um you know more credibility because he's been in the system and now speaking out against it we have we haven't been able to treat data in a correct fashion in the last 10 years and Scandals like Cambridge analytica show their app to everyone and every in every part of the world so asking for a six-month stop sounds more like give me a little bit uh more time to come up with my solution more than let's find out the correct solution here in Europe the artificial intelligence act that is quite well made I I must say has been in the works for the last five years I think and there are still things that sort of evade the kind of barriers that are set in places so stopping and asking for a six-month uh sort of truce between humanity and artificial intelligence isn't really going to do much more than a PR stand and in PR stands Elon Musk is quite good all right but you know we must mention the concerns about social media influence and politics Society was very sharp in the aftermath of the 2016 U.S presidential elections right we had all those accusations about voter manipulation by Foreign agents so there are many concerns but if I can ask you an ankush sabar while you were smiling over there so let me ask you what is because some of these the words in this letter can be quite alarming right so what would you be able to just put together for us uh for our viewers what the worst case scenario in terms of AI being quote catastrophic to humanity could be see before that you know I agree with Matthew so uh see I respect uh Elon Musk whatever he has done and his purpose is of course serving Society with that but he's a great marketeer uh right so you see one of the people who have signed this letter is the co-founder the founder CEO of stability EI which is Fun by Elon Musk right so I'll not talk much much about it but see I think it has in fact I think someone said when um calculators were coming and you see a lot of social media we're getting a lot of memes where a lot of mathematicians said hey let's stop this give the moratorium the calculator should not be there and should not be right so everything comes away from dance and see what are the Hypes and whatever is the reality of chatbot if that is understood and that that should be given it as a concrete uh disclaimer I think the worrying part about the GPD is so they are ready to answer you everything right so they are not saying hey you can ask me only about this subject this subject is saying hey you asked me anything and they say everything with so much confidence and without any disclaimer whether this can be wrong right so we have to work towards the you know sanitizing and alarming the users okay I want to just quickly ask you all about something that stood out to for me when I was trying to do some research on this this speculatively whether letter r should we develop non-human Minds that might eventually outnumber outsmart make Obsolete and replace us um can I just get your responses to that kazim rizwi does in the future is AI going to pose a more General threat to human control over our own civilization is that like you we see those things in the movies no I I don't think as possible because I think basically what what AI is right now and what we see AI is the ability to understand and detect patterns uh by collecting processing large volumes of data right that's the fundamental with respect to how AI performs it does not have any human any emotional intelligence it does not have uh empathy it it also does not have any type of Consciousness be a lot of human consciousness so I I do not think at least at this stage and even going forward that AI would be able to replace human binds and then bring itself together and then you know same question to you this this whole talk about Singularity right what is that it's a hypothetical future where technology surpasses human intelligence and change changes the path of our Evolution some are predicting it's going to happen by 2045 your your opinion on this so the existential risk as you mentioned is a good diversion from the current problems that we have with AI like uh like people like casimanda pointed out the bias on few other things uh now coming to the question artificial super intelligence or artificial general intelligence the human level I think it's won't be a existential risk because we tend to compare a human intelligence with dominance so we think we as humans we are the most intelligent things on the planet and we ended up dominating this planet so we think something that would be smarter than us end up dominating us but if you look at a lot of that has to do with our dominating Nature has to do with how uh the natural level natural selection work Evolution worked and all those things so that's there but uh to be honest we are nowhere close to having a human uh human level intelligence and it will take maybe beyond our 2045 or more than that but I don't think there will be an existential threat from these models of more or these kind of AI like the way we think you know uh Terminator kind of scenario that's not gonna happen because intelligence is not always equal to dominance I mean we have seen female elephants which is very intelligent but they are not a dominant uh way they are very Cooperative elephants are in general so I think these kind of uh narration should change and we should look at the real problem that's happening right now the job displacement uh the bias the security the Privacy these kind of problems are most importantly how these models are built on top of Labor from uh the Dual countries right people in Kenya were annotating and working and building this data behind HR GPT and what at what two dollars per hour so this kind of Labor exploitation also should be considered all right do I have 30 seconds I can quickly ask Mateo Flora your response to on this whole uh the fear that it will surpass human intelligence it's independent thinking no longer needs to take orders from humans it takes on a life of its own we are very very very very very slow in incrementing our intelligence and our level of intelligence has been constantly stayed easy for the last few Millennia the first artificial intelligence to be able to to show some kind of intelligence will overpower Us in the space of a few seconds but if you want to take a much better look than me at this kind of uh issues you you better take a look at Nick bostrom's book super intelligence and he has been talking about that for the last 10 years I think all right well thank you all for being part of this conversation for being on We the People it's an important issue uh it's been making headlines not enough shows were done uh on it this week A lot of people are curious about it so we're grateful for your time for trying to help separate uh the noise from the facts thank you for joining us foreign

Channel: NDTV

Views: 8,323

Keywords: NDTV 24x7, breaking news, latest news, live news, ndtv, ndtv live 24x7 english news, top news, Artificial Intelligence, AI, Tesla CEO Elon Musk, Artificial Intelligence threat, Artificial Intelligence research, silicon valley, development of Artificial Intelligence, training of Artificial Intelligence suspended, We The People, Sarah Jacob, TV Show, Show= We The People, Anchor= Sarah Jacob

Id: PZT0A5xu4OE

Length: 26min 56sec (1616 seconds)

Published: Sun Apr 02 2023