AI is a Lie - Cutting Through the Hype

AI is everywhere and everyone is talking about it but as it turns out the vast majority of what they're saying is misleading at best and at worst an outright deception while the feature on your Smartwatch or your new co-pilot PC is called AI it's not the AI that you're probably thinking of so we're going to take a reality Hammer to this hype machine and break down what AI can do what it can't do and why the tech industry is so eager to slap the AI label on absolutely everything and guys this Rabbit Hole goes way deeper than you think okay predictably it ends at money but the path to get there it turns out is really confusing on purpose which makes it very interesting like this interesting message from our sponsor MSI their AIS rs2 is a pre-built PC touting some impressive specs like an Intel Core I74 1400 KF 16 gigs of ddr5 RAM and a GeForce RTX 4070 super check it out in the description the classic definition of AI is probably best Illustrated with fictional examples it's what you see in sci-fi Creations like Commander Data Hell 9000 and GLaDOS these are are computers or machines that demonstrate a capacity for reason however naive twisted or alien it might seem to us meatbags now you'd be forgiven for thinking that that's still the definition of AI a lot of people seem to think that it is but in reality the meaning of words is ever shifting and we would now refer to these characters as having a gii or artificial general intelligence what you're referring to as AI then is in fact narrow AI or as I've taken to calling it AI Ani is not a general intelligence unto itself but rather another component of a fully functioning system made useful by specialized algorithms and data processing utilities forming a complete artificial intelligence system didn't think I could make that point in the style of Richard stallman's famous interjection well I could haha but I also didn't have to that previous paragraph was actually written by gp4 Omni and this is exactly the sort of thing that modern AI does very well and that's because most of the time when we hear the term AI we're actually referring to machine learning a subset of AI involving algorithms that can analyze patterns in data they get trained on things like text multimedia or even just raw number outputs and using this training data they identify patterns through statistical probability they can be further trained through reinforcement learning then by rewarding correct outputs and punishing incorrect outputs kind of like training a hamster the results allow these algorithms to summarize predict or even generate something seemingly new and in many cases they are so impressive that a good machine learning system can be indistinguishable from classic AI or a gii well then minus if it looks like an AI and it quacks like in AI what's the difference well artificial narrow intelligence is limited to specialized tasks gp4 Omni specifically is a large language model which means that it is trained to understand and generate natural language like the words I'm speaking now it's basically an autocomplete on steroids what sets it apart from your phone's keyboard though is that it can also process information based on patterns that are learned during training including definitions mathem iCal formula and so on and so forth that makes it capable of generating unique output that wasn't part of its training data GPT has traditionally been incapable of image video or audio generation there are other types of generative models like Sora sunno or dolly that feature their own specific talents but most of them are incapable of operating outside of their specific Niche and all of them are limited by their training data in a similar Manner and because they're limited by their training data in many cases the answers that they give resemble their training data which if you're an artist or a photographer and your work gets added to a model is probably not your idea of fair use much less a good time worse when generative models are faced with a concept that they don't understand or they simply run out of tokens they can begin to hallucinate that is to say they just make things up as they go which is why sometimes you get eldrich Abominations like these with that said these limitations don't mean that machine learning AI is a dead end it's been deployed very effectively for diagnosing diseases and in other highly complex scenarios where the data is dense and the conclusions require interpretation these specialized models are extremely useful they're just also extremely not new simple neural networks have been in use for decades for things ranging from handwriting recognition to web traffic analysis and yes even video game these are not scripted sequences the AI is determining when your allies choose to advance and how to best help you out in combat and chatbots the main difference is that they run much faster on Modern Hardware if I had to distill down what artificial narrow intelligence really means then I would say it's like having a thousand monkeys at a thousand typewriters with a thousand pieces of reference material for what the outputs are supposed to look like with enough trial and error then they do arrive arrive at a point where they're likely to spit out a correct or at least correct enough solution then we take all those monkeys and we take a snapshot of the model State and we start feeding it inputs for both Fun and Profit what ani is to a brain then is kind of what a single app is to a computer it's a building block it's something your brain is capable of but it's just one of its many many functions shifting GE a bit then what would artificial general intelligence look like well it would need to be able to handle everything we've talked about so far just like your brain can take some past experiences and turn them into a new creation but again like your own brain it would need to be able to run many of these models concurrently and continuously train and iterate on them rather than relying on fixed snapshots only then would an AGI have the ability to truly learn and adapt to new things bringing it closer to that that classical definition of AI and really blur the lines between machine learning and machine Consciousness the problem is even if we had software that sophisticated we are nowhere close to being able to run an AGI even on a modern supercomput let alone on your AI smartphone but all right lonus you still haven't explained why any of this is even a problem I mean freerange meat is just marketing bollocks too so who cares well well truthfully in most cases I don't I mean cooler Master's AI thermal paste snafu I was never bothered by it because I never expected my paste to be sentient anyway but there are situations where this kind of marketing can have an impact on user safety and therefore does matter let's talk about Tesla Mr musk has said among other things that any vehicle from 2019 onward will be able to reach full autonomy and he's certainly put out some impressive demos both canned and even in the form of public beta software that you really can use and that's really cool but unfortunately it isn't much more than that you see to operate a vehicle safely it's not enough to be trained with images of painted lines and traffic cones stop signs pedestrians vehicle Telemetry data it's not even enough to be trained to predict the likely maneuvers of nearby vehicles and life forms on the road anything can happen and by definition by its very definition Ani is not capable of handling an edge case that it has never seen before even if it was by the way I have some really bad news for you Tesla owners out there Hardware 3.0 has about 144 tops or trillion operations per second worth of processing power for context Windows 11 recall a feature that does little more than take screenshots and analyze your PC usage for search asks for 40 tops now to be clear tops is not a be all andall measure of performance and there is no way that Microsoft has optimized the code for recall nearly as much as Tesla has for full self-driving but this should still illustrate the point that Tesla either did or should have known that a vehicle with the AI capabilities of a family of iPhone 15 Pro users would never achieve that kind of realtime contextual awareness that's required for complex situations like operating a motor vehicle and they misrepresented its capabilities in order to sell more software that was never going to leave beta that is going to be a doozy of a class action and it's a common story that has led to this current mess where fuzzy definitions and impossible promises have turned AI into this meaningless buzzword like all the rest of them all of them refer to legitimate useful Technologies some of which have really come to fruition but their meanings have become diluted with overuse and it means that when computer cognition finally happens we're going to have to call it something completely different in order to differentiate it from all of the marketing wank on the subject of marketing calling ai ai wasn't an accident the people behind that marketing know what you think AI means they know that the promise and the Mystique behind the term is tantalizing and they know you'll click on an article that is interviewing an AI expert who discusses how dangerous or already alive it is they want you to buy into their hype they want you to buy into their stocks what we have now though really are decent summarization engines and lukewarm guessing machines that are tuned for working with different types of media they can't reason and anyone who's seen chat GPT get something ridiculously confidently wrong can attest to that and they also can't understand why they get these things wrong which mean they can't learn or improve on their own even if you explicitly tell them they also have a finite number of things that they can remember called tokens which limit their ability to maintain a train of thought for very long anyone who's tried to get chat GPT to write a novel or even a moderately complex python script can probably attest to that of course as time goes on some of these limitations will begin to lift and these a Ani models will start to look more and more like a GI to the lay person I mean just ask the guy who had Bing fall in love with him but at the end of the day it's still the same thing as it's always been to paraphrase a paper co-written by Emily bender and Alexander caller it's a hyper intelligent octopus that is observing and learning the patterns that are expected of communication and repeating them back more accurately over time so when faced with a novel situation like a bear attack that octopus has no hope of being able to guide you on how to defend yourself with the materials you have on hand it has no concept of bear or stick only what words most closely match the pattern that it has previously observed to further illustrate this if you haven't tried it before try to craft a prompt that coaxes stable diffusion to spit out something very specific it turns out it's pretty challenging and when it finally does it it's generally composed of a mosaic of shapes and patterns from its training model that best match the keywords that sounds overly simplified but that's because it is in essence what it's doing it takes the keywords from your prompt and starts compositing filling in the image until it hits for example a certain percentage of computer and desk and it does this without any context so if you asked it to show you a computer would it show you a desktop PC a laptop a server rack and for that matter from what era the answer is often simply yes because it lacks the context behind what computer means to you and it can't always even for a coherent depiction even in the same image so then Linus just don't worry about it and wait till AGI shows up and you don't have to bother with online dating anymore well not quite as generative AI models learn to create works that are closer and closer to our expectations and to finally understand how many fingers and teeth humans normally have we're going to start to find it more and more difficult to distinguish their output from real photographs or artwork or other people ushering in an era of unprecedented distrust and the bad news is that the folks in charge of helping us deal with the consequences of all of this have a lot less funding than the ones who are trying to sell it to us so be safe out there and safely check out our sponsor manscaped if you're looking for a way to quickly take care of your more delicate Parts manscaped is on the case with a performance package 5.0 Ultra it includes the lawnmower 5.0 Ultra to trim the grass as well as the weed whacker 2.0 for your ears and nose so it doesn't feel like you're always tickling yourself you'll even get their crop Soother made with shea butter and Coco seed butter designed to pamper your delicate areas with essential moisture and soothing relief and for an added layer of freshness give your package a Royal Treatment with their crop preserver to keep the funk at Bay check out the performance package 5.0 Ultra at manscape.com techtips and use code techtips for 20% off and free shipping if you guys enjoyed this video and you're looking for some more debunking spelunking maybe check out recent video on PC speed up tools to see if they really do what they say they do

Channel: Linus Tech Tips

Views: 1,652,252

Keywords: ai, artificial intelligence, machine learning, deep learning, ml, dl, computer intelligence, synthetic intelligence, openai, chatgpt, xai, google, microsoft, copilot, bard, mistral, claude, gemini, GPT, GPT-4, GPT-4o, anthropic, apple intelligence, compute, TOPS, GPU, smartphone, ANI, AGI, artificial narrow intelligence, artificial general intelligence, weak AI, strong AI, generative AI, Tesla, Full Self Driving, FSD, Stable Diffusion, DALL-E, Sora, Suno

Id: nGIpdiQrFDU

Length: 14min 35sec (875 seconds)

Published: Thu Jun 13 2024