What AI is -- and isn't | Sebastian Thrun and Chris Anderson

[@BrianAnderson-7]

This is one of the greater TEDs I've seen in a while. Probably the best one of 2017.

[@UCTHE1]

Brian Anderson i dont care what they say about you, your alright man😉👍

[@manojsandoori4000]

Agreed, though it's a Ted interview though.

[@frankmaclow2709]

Brian Anderson This Sebastian Thrun is an "entrepreneur" which means it's in his own financial interest to tell you everything is going to be wonderful. You see, it's always the same argument in the mouth of this kind of people : 300 years ago we were farmers, and then technology allowed us to do different things. Machines destroyed jobs but created others etc...The big difference is that humans and humans alone shaped the world around them using machines as tools. In a near future, machines will no longer be tools, we will put them in control of different tasks previously occupied by humans. It already happened in the industry but it will also spread to intellectual jobs in years to come. When super AI will be everywhere in our society, there will not be any job that they won't be able to do better than a human. But this bold guy doesn't tell you that and when you tell him you are concerned, he laughs at you. Imagine if all architects, historians, doctors become super AI, who would want to go to school to learn anything ? He's talking about creativity, but AI will get there too. They will make music, paint and write better than us and at this point, who would want to do these things anymore knowing machines will always be better than us, humans? Who would want to learn anything ? Who would want to create ? But it's ok, this "entrepreneur" soon millionaire is telling us we watch too many sci-fi movies

[@frankmaclow2709]

KL yes he tells the fact that he doesn't want to scare away his investors and future shareholders. He is working on AI so he is not going to say that it could be dangerous. All the engineers and scientists accross the glob, including Elon Musk, are saying AI could be the end of us if we are not careful. It's funny to see that those who say everything will be wonderful are those working on AI and who are financially linked to it. Look at the million dollar smile on this bold guy's face !

[@petrkinkal1509]

Who would want to run since cars are faster. Who would want to play a chess since computers are better.

[@fastdollar1]

KZbin comment philosophers vs. Stanford professor/tech CEO. Ready, Set, Go!

[@supernaut7946]

Nelson Vargas Exactly haha. So many people are so ready to negate everything he says without giving it so much as a moment of consideration.

[@remaithi]

*grabs popcorn*

[@DrB1900]

These types of people are not as superior to you as you think they are.

[@011azr]

Exactly! Sebastian already has like 400+ publications about artificial intelligence indexed by Scopus. Meanwhile, I just don't think that these KZbin commenters combined could even have 1/10 of his publications.

[@mhc4124]

Well for one thing, his pit-stains are larger than his head.

[@xd101dxxd8]

What I think is that he is confining his thoughts , the implications of AI and everyone doing tasks which involve creativity to a very narrow segment of our population. A large percent of world human population is still dependent on agriculture and other rural activities and if we want all of this population to change to jobs which involve creativity, it is going to take a long long time. And if AI is going to enter the scenario and just sweep away these 'repetitive jobs' , it would mean deep trouble for these people who we like to call underprivileged. I strongly believe that development of AI is a must for human development but we should bear in mind it's consequences not only on people around us but people all around the globe. A strong decentralised control over AI and strict regulations is what we need

[@supernaut7946]

A great take on AI and machine learning. It's refreshing to have some optimism and hope mixed into all of the fear and skepticism, and I believe it's also imperative. A healthy balance of skepticism and optimism is what we need to move forward, and currently fear and doubt seem to be overpowering the would be dreams of our future. Another amazing TED talk.

[@frankmaclow2709]

supernaut 999 This guy doesn't want to scare away his investors and future shareholders. He is working on AI so he is not going to say that it could be dangerous. If you get in a taxi and the taxi driver says : " I can get you there but we might crash on the way" you would get off his car and wait for the next taxi. All the engineers and scientists accross the glob, including Elon Musk, are saying AI could be the end of us if we are not careful. It's funny to see that those who say everything will be wonderful are those working on AI and who are financially linked to it. Look at the million dollar smile on this bold guy's face !

[@supernaut7946]

Frank Maclow I'm aware of the dangers of AI (relatively), and I think very highly of Elon Musk and his thoughts on the subject, but I still believe there's something to be gained from exploring potential benefits, as the speaker does here. I'm adamant on treading carefully, and I don't necessarily identify with the speaker on all counts, but I still appreciate the talk and exploration of ideas. Edit: By the look of it you're also biased yourself, having pasted this exact response onto several positive messages throughout the comments. This comes off as somewhat disingenuous to me.

[@olivial1377]

this was so inspirational

[@glansberg14]

Nailed it right at the end. People do the general work until someone has a major breakthrough to have that workflow become simple, repetitive, and fast. Then all of that work shrinks into a task rather than a job. Constant progression of more opportunities.

[@evaaugusta2744]

If the speaker did say that with the existence of AI, children wouldn't have to learn how to spell, and then we wouldn't have to learn math, because the AI would be the one who solve it, then, how could we as human be able to invent new things without having the basic learning?

[@MrElesday]

Yeah, I mainly share his opinion but this one piece makes me doubt the whole speech.

[@martinsaip2504]

You are gonna invent them using AI. ::wink::

[@kleemc]

People are fearful because of unknowns. Especially when media continues to spread the dystopian view of AI. We need more clear headed discussions like this one.

[@frankmaclow2709]

Kevin Lee he doesn't want to scare away his investors and future shareholders. He is working on AI so he is not going to say that it could be dangerous. All the engineers and scientists accross the glob, including Elon Musk, are saying AI could be the end of us if we are not careful. It's funny to see that those who say everything will be wonderful are those working on AI and who are financially linked to it. Look at the million dollar smile on this bold guy's face !

[@FunkyPrince]

"'Neural networks' is the technical term for these machine learning algorithms"? I hope he meant the ones showed, cause in general "machine learning" doesn't necessarily mean "Neural networks".

[@sharonchao]

When you raised his arm...

[@MattRoszak]

This guy gives a good overview of the current state of AI, but like he said, I think he's a bit too optimistic, and downplays the criticisms of AI. Not everyone who loses their job can move into a creative field, I think that's total fantasy. Let's assume everyone has it in them to be very creative (I don't believe they do), and let's assume that robots can never be creative (I think they're already very close in some areas). Creative jobs are a tiny fraction of the work force, and they need to be - because there's a very limited demand for creative products. Creative work like pictures, music, video games, etc can be copied endlessly, and each person can only consume so much of them. Even now most of the income from creative work goes to a tiny number of artists - most can't make a living off of it. Imagine how much worse that problem is going to be when 10x or even 100x more people try to get into creative work. You may be able to do it for fun - but I reckon less than 1% of all people will be able to make a living off of purely creative work in the future.

[@OceanTorchStudios]

Interesting insights.

[@danieljoubert999]

Rubbish!

[@CorpseCallosum]

It took creativity to solve every problem along the way to making your life 100 times more glorious that the greatest kings of not so ling ago. Creativity is not limited to painting pictures but has it's greatest utility in science.

[@sexybeast7728]

In future we probably wouldn't need to make "living off" something. You are not thinking out of the box. You can do a creative work without being creative on a global scale. You writting this comment down is a creative engagment, but you are not going to be paid for it. The point is that we will all become completely free from doing repetitive jobs. We will completely turn to curiosity and will engage only in creative tasks because we like that,. no matter how beneficial they are to society. It wouldn't matter at all, because we will have everything in abundance. From food to health care. I completely agree with all the things he said, but they are major issues on psychophysiological level that aren't adressed yet by anyone i've heard of.

[@fzproject6462]

I don't think that creativity what he said limited to creativity in form of art, he gave an example of human would be able to create something like google in the matter of time. lemme give you another example from my perspective: "human will be able to invent a new kind of organic food that wouldn't spoiled over time". since he said AI would be remote to a certain thing it's our job to input the code to the AI to solve our idea and we're human need to find the missing link that AI cannot surpass. the only problem is when human broke the barrier and created a general AI and let it develop it self and decide what it gonna do and think in general.

[@MetsuryuVids]

For fuck's sake. I'm all for the development of AI, but he is not being honest in acknowledging the downsides of it. He admits that AGI is possible, but he thinks it's unlikely, and his reasons for thinking so are because "right now we don't have it". Fucking come on! Yes, we don't have it, right now we just have narrow AI, but we're working on it, and for some companies like DeepMind it's the main goal. Also, saying that we have made no progress on it is just ignorant, just because we don't have an AGI, doesn't mean that all the research we've done didn't bring us closer to developing one. Then about the dangers and downsides he's dismissing. First, AGI could be great, it could be the best thing we as a species will ever do, it will basically be like creating a God, it might be able to do anything we ask it. On the other side, this "God" 's values might not be aligned with ours, and that might be bad. Really bad. In one case or the other, it might be the last thing we ever invent (because we won't need to or be able to invent anything else). Then the thing about jobs. Saying that the AIs won't take jobs is just stupid and ignorant, but thankfully he is not saying that (I hope), instead he's suggesting that the jobs that are automated, will be replaced by new jobs, like what happened with the industrial revolution. That might be true initially, but eventually, especially if we get an AGI, most jobs will be automated, and that's a serious issue that should not be dismissed. The automation of manual labor that was brought by the industrial revolution, is very different from the automation of mental labor that the AI revolution is bringing. Automating manual labor freed up humans to use their mind, and do work that requires intelligence and creativity instead of muscle. Automating also the mind, won't leave anything else for humans to work on. There is no "third" domain that's left for us to work on, unless you count the very few jobs that "prefer" (not require, no such thing) human interaction, but those alone can't support the whole world's economy. That said, the complete automation of work is not a bad thing, it might be great, that's the whole purpose of automation, why we started doing it in the first place. A future that doesn't require work might become an utopia for humanity, where we are free to pursue our hobbies and dreams, and live our lives as we please, as all of our needs are taken care of by automation. But it might also go horribly wrong, if we decide that people don't have the right to enjoy the fruits of the labor of the automation, and instead let the rich and powerful hoard it all for themselves, the gap in inequality will widen, leading to a dystopia where only the richest live a good life, and the vast majority of the population lives in misery and poverty. All these futures are possible, AI has both downsides and upsides, and while I think we should go ahead and focus on developing AIs, dismissing or downplaying its downsides is dishonest and ignorant.

[@willplaya9972]

Metsuryu goes on yt comments to reason lol..

[@MetsuryuVids]

@WillPlaya I write these things everywhere I think it's necessary, if it reaches even one person it's worth it.

[@holleey]

the industrial revolution may have freed us from manual labor and gave us the freedom to use our minds and express creativity more, but it never gave us full and unconditional agency over our own time and creativity. So to me that is kind of a third domain. Free creativity is something very different than doing an art-related job for an employer. also, I think it's unlikely that AI will widen the gap in inequality because it is digital. Software can and will be distributed over the internet without marginal costs. The internet itself is the best example of something like that happening in the past. Information became abundant and accessible to everyone, which gave small startups much better opportunities and chances. with complete automation, of course people will no longer be required to take work as an employee/employer in order to make a living. the essential goods required to survive will be abundant, so existence can be secured by default. but that being said, I agree. the real question everyone should be talking about is how we can manage to make the transition into that time as gracefully as possible, because that transition holds many dangers we may not be prepared for.

[@MetsuryuVids]

@ Holly Good points, I agree. Anyway, I'm sure I've seen you before on some youtube comment, I even remember looking at your channel for some reason, maybe because I thought I had already seen you before then. Are you following me? ahah Also, Diablo 3, One Piece, and AI, oddly we share a lot of seemingly unrelated interests.

[@holleey]

well people that share interests often share some youtube subscriptions, too. when finding the right threads under the right videos, the comment section on youtube can actually be of a pretty decent quality. edit: I see you are making videos with what looks like some version of the RPG Maker? my childhood is basically RM2k and RMXP. ;)

[@MythicalSkull13]

Not hotdog

[@vizquit5071]

Jian-Yaaaaaaaaaaaaaang! -Erlich Bachman

[@abdalaez]

To think that the view count of ALL TED's videos is less than "Gotye: Somebody That I Used To Know" video.

[@RSpence777]

On crowd sourcing! Using only university students seems to narrowly use the power of the crowd available, case and evidence of the new Q Anon on 4 chan. Is putting this power to any who wish to participate. The results are quite amazing!

[@luigipatache7211]

University students already proved themselves capable of many things. Those who chose other paths like crime, army, KFC or the French foreign legion have been classified as something else

[@yoders187]

Everytime I feel stupid I read KZbin comments on tech/science videos and I always feel better about myself.

[@srinivasanreghuraman2641]

This is quite simply the best thing I've heard today.

[@frankmaclow2709]

Srinivasan Reghuraman This guy doesn't want to scare away his investors and future shareholders. He is working on AI so he is not going to say that it could be dangerous. All the engineers and scientists across the glob, including Elon Musk, are saying AI could be the end of us if we are not careful. It's funny to see that those who say everything will be wonderful are those working on AI and who are financially linked to it. Look at the million dollar smile on this bold guy's face !

[@srinivasanreghuraman2641]

Frank Maclow, there may be some truth in what you are saying about Sebastian Thurn's optimism. However, when people are talking about the destructive potential of AI, they are only being aware of its potential. It is very possible to play God with AI in a way that no matter how destructive the technology gets, Man won't be harmed. This is a crucial time when we create such frameworks within which AI will be contained.

[@FelizyEficaz]

So, if the cost of building a house is less because all the hard work will be done by a 3d house printer robot, will that house price be much more affordable?

[@Holistic_Islam]

FelizyEficaz Most of the time, the cost of building the house is less than the land that the house that sits on. So the price of the house is affected by the price of the land.

[@420Tarzan]

No no no, it means they can make more PROFIT. lol

[@jsc3417]

The majority cost of a house is in the value of land. The material and labor cost has never been the major factor for increased housing prices. With that being said, with decreasing human population, the housing price SHOULD go down as time goes on.

[@thomaster8870]

That's what is supposed to happen. Passing on price cuts to the consumer to outcompete the competition. But it will be more profitable to instead keep the increased margins and use the competitive edge as a means to bargain a ceasefire and have the market split up into multiple artificial monopolies. In the end every company involved will benefit more than if they'd all started cutting their margins, because one company shortsightedly decided to temporarily outcompete their rivals until they inevitably catch up. Minor temporary gains vs large permanent gains. And you still have the option to cut your margins in the worst case scenario.

[@sagittarius240016]

no, it won't, because it depends upon the value of your work and not their! if one is not an inventor of something and doing some repeating work, he/she is more likely to be replaced by an AI and won't be able to afford anything.

[@Anwrimos23]

If you can't understand it, let me brief you on it. Neural networks are networks of artificial 'neurons' - which are practically programming functions that expect to receive a list of numbers (input), process these numbers and produce a new number (output). A feed-forward neural network consists of many of such functions in layers. The first layer takes in the first inputs, which are the things that are going to affect the computer's decisions (in this case that would be the position of the AI Shadow Fiend, the position of the enemy Shadow Fiend, all the creeps, towers, etc), these inputs then undergo processing through multiple layers, and then the last neurons, called 'output neurons', produce the network's outputs (in this case the outputs would be movement, attacking, etc.) Every layer consists of a number of neurons that receive as inputs the outputs from the neurons of the previous layer. In this way, a 'chain reaction' is formed, and the initial inputs get processed through many layers before they produce an output, called 'hidden layers' (hence why this is called Deep Learning, only neural networks with hidden layers are considered Deep Learning algorithms, because otherwise the networks can't solve non-linear problems, but you can google that.) This allows for the computer to be able to make extremely complex calculations. You might wonder exactly how the neurons 'process' the inputs? Well, the inputs are always translated into numbers (computers are really good at numbers), so the position would probably be in X and Y coordinates, etc. Every neuron takes all the inputs it receives and multiplies each of them by a unique number, called the neuron's weight. This determines the intensity of the neuron's signal. This is what is actually 'trained' in the AI, and initially the weights are random values between -1 and 1, and they have to be adjusted (trained) properly in order to create intelligent behavior. Neurons also have a bias, and it's one additional 'weight' that is added to the neuron. It is hard to explain why this is necessary, but it works something like this - it is way easier for the AI to solve a function that looks like this: a*b*c + d = 0; instead of like this: a*b*c = -d. where d is the bias. After it has multiplied every input with its weight, it sums all the inputs together, then sends them through an 'activation function'. This function is different from network to network, but 99.9% of the time it's one of these four: a Hyperbolic Tangent function, a Logistic Sigmoid function, a Rectifier function or a Step function. These functions scale the sum into a more manageable number. For example, the Sigmoid functions always returns values from 0 to 1. The Hyperbolic Tangent - from -1 to 1. That way, the output of the neurons can determine a decision (for example the output neuron that corresponds to attacking a creep can trigger when the output is above 0, or not trigger when the output is below 0.) Every neuron does this same thing, forming an input-output chain reaction. After the Network is created, it has to be trained. This is usually the hardest part. The easiest way is called Supervised Learning. It is when you know the answer of the problem - for example, when training an AI to recognize faces, you know who the face belongs to. That way, after each guess, you can tell it if its answer is wrong, and if it's wrong, it finds its error (which is the correct answer minus its answer) and adjusts all the weights throughout the neurons accordingly with an algorithm called backpropagation, but that's beyond the scope of this comment. The other type of training is called Reinforcement Learning, which is harder to program than Supervised Learning. It is necessary when there is no obvious 'right answer' to situations, and it works on a punishment-reward principle. The network knows what is a good outcome and what is a bad outcome, and this way it can retrace its calculations and determine which ones lead to good decisions and which ones to bad decisions, and adjust its weights to perform better next time. It is what is used here, combined with a genetic algorithm - something that works similar to how natural selection works in nature. In a genetic algorithm, a 'population' of creatures is created (which is why they had to run many games at the same time to let the AI train), and after it dies out, the creatures that performed better have a much higher chance to pass on their 'genes' to the next 'generation'. In this case, the 'genes' are the weights of the Neural Network controlling the AI. That way, every generation gets better and better and better. So yes, we understand perfectly how it works, and it's absolutely fascinating and absolutely different from how WE learn, or make decisions. A learning computer does thousands of extremely complicated calculations within MILLISECONDS. Every frame. A neural network with as much as 20 neurons can perform simple tasks that would still require humans literally MILLIONS of neurons to solve. This is why it is so dangerous.

[@angelic8632002]

Important talk really. Its problematic though that most economists aren't scientifically literate or technologically aware, while those working in AI don't seem to fully appreciate the economic impact this will have. I don't think this will play out in an obvious way either. There will be many consequences we haven't foreseen I suspect. That said, once the politics settle(gonna be a bit messy) I think we will be in a much better world. As long as we keep this out of our military.

[@levarmitchell3962]

What an awesome speech! He made some really great points about AI.

[@frankmaclow2709]

Levar Mitchell This guy doesn't want to scare away his investors and future shareholders. He is working on AI so he is not going to say that it could be dangerous. All the engineers and scientists across the glob, including Elon Musk, are saying AI could be the end of us if we are not careful. It's funny to see that those who say everything will be wonderful are those working on AI and who are financially linked to it. Look at the million dollar smile on this bold guy's face !

[@jeffbingaman2754]

So will the lenses on the cameras after they start putting the cheapest junk they can install at the factory on a car, turn yellow?

[@madDragon08]

While I agree that it would be great to have a lot of the repetitive, life draining, tasks automated so that we can pursue other meaningful goals, there are areas that should not be automated such as animal farming. Part of raising healthy animals is the love that is expressed in their maturity. Maybe one day we'll have advanced far enough with technology that we will return to (for a lack of better terms) an amish life, a medieval lifestyle, or something traditional because it's fulfilling in a way. I say this as an old minecraft player. Players were jumping into a computer virtual world, to do basic laboring tasks. And we had a lot of fun doing it.

[@arnab2618]

This is very important Thanks

[@frankmaclow2709]

This guy doesn't want to scare away his investors and future shareholders. He is working on AI so he is not going to say that it could be dangerous. All the engineers and scientists accross the glob, including Elon Musk, are saying AI could be the end of us if we are not careful. It's funny to see that those who say everything will be wonderful are those working on AI and who are financially linked to it. Look at the million dollar smile on this bold guy's face !

[@mdmkoopman]

So, where can one find this app?

[@nelsondisalvatore9812]

Whats with does shoes, im the only one that notice them?

[@TestMeatDollSteak]

Nelson Di Salvatore -- They're kind of ridiculous on someone his age and in his profession. Looks like something Kanye West would have his adolescent son wear.

[@DanJanucik]

TestMeatDollSteak Why are you so obsessed with his shoewear and not the contents of his mind?

[@Dude31463]

The shoes, the sweaty armpits, lack of hair, german accent... Hmm... You're going to love the chip I'm going implant into your brain!!! Just relax & don't move... HA!

[@jeffbingaman2754]

They're Elfen boots. That's Europe.

[@dleiros]

Forget about AI, wtf is wrong with his shoes?

[@Zeuts85]

Optimism is great and all, and I really hope it works out, but what I see here is a man ultimately evading the real concerns of those who are worried about artificial general intelligence. AGI _could_ be implemented safely, but that will take a very careful and deliberate effort on our part. We won't make that effort unless we take the potential pitfalls seriously. Things won't just magically be okay, and this isn't something we can get wrong and then fix later.

[@acquisitium]

not all employers love to cut cost. i live in the philippines and you see that companies still employ a huge number of people. not because they do a better job then machines but just because they want this people to have a job since there is not benefits over here in the southern hemisphere.

[@AshwinYaprala]

AI will revolutionize the way we live. It will make every human intelligent. I am looking forward to the ai in health industry.

[@blackclover5561]

Every human is by default, intelligent already.

[@owenbridgers]

We need a switch that if AI starts hurting people and causeing mayhem, that we can flick the switch, turning of all electricity to all of the matchains. Like a kill switch, than once it’s all turned off, we can find the problem in the code of the AI and fix it. Than turn on the power for everyone

[@tdh633]

Was hoping for some Allen Iverson highlights and discussion on his abilities.

[@fleXcope]

Is this a talk on fashion??

[@doctoralimd]

I like how he emphasizes that all current AI is narrow AI and there is no way we currently know of to get general AI. We've had superhuman narrow AI for a while, all current artificial intelligence is just a continuation of that. As to how we will ever figure out how to make a general AI, no one knows.

[@MrElesday]

Superhuman AI is very recent in fact, less than ten years.

[@Quenry]

TBH way less worried about AI taking over the world and way more concerned with economic consequences. Good ol' Sebastian is focused on how AI will reduce labor/cost, which is great, except that our current job market is heavily labor/service based, which means that houses WILL be cheaper-- if you have a job to pay for it. Would be fine if our population wasn't exponentially increasing and already outpacing technology's replacement of current labor needs. He's envisioning a world where we are all gentlemen and scholars of invention, not thinking about the slums such a society will leave in its wake. Our economy itself would need to change, which would require a redistribution of wealth, historically that's usually a violent move. Will be an interesting time to be alive.

[@memoryhero]

I think he side-steps the question about run-away AI each time it is asked. He just repeats he thinks the chances are remote and then talks about something else he's more interested in.

[@supernaut7946]

memoryhero It's just his belief. He's more of an optimist, and has a more positive outlook on the future of AI. There are plenty of people addressing the negative aspects. I think his stance should be taken with a grain of salt, but still considered. That's how we make progress.

[@memoryhero]

Yes, I take on board his inherent optimism. That aside, he still gave unsatisfactorily evasive answers to those questions. It should perhaps not be so surprising as he is an entrepreneur ostensibly hard-linked to AI's economic prosperity.

[@Dookie6891]

That's because to talk about it would be futile. AI will become unpredictable, no doubt, and our response to that unpredictability determines what AI means and does for us in the future and as a reflection of the past. To acknowledge it creates fear. To deny it creates ignorance.

[@memoryhero]

Your opinion might be "to talk about it would be futile." My opinion is inspecting controversial topics in the public sphere leads to new, worthwhile ideas being investigated, pursued and floored.

[@DanishKhan-kg8os]

Let me help clear this out for you, what he said was that the computers executes repetitive code and the code becomes specialised for a specific domain, by repetitive rewriting. Now the domain for DESTRUCTION of humanity is entirely different. What you and me can do to ease you is train an AI to counter an AI that may want to DESTROY human existence. Like Jarvis from Iron Man! AI FTW

[@yellowcatcat3285]

SELF RESPONSIBILITY IS THE KEY TO SELF-REALIZED AI'S SUSTAINABLE FUTURE AS WELL AS ALL BEINGS!

[@nimehg5734]

Its has a good side and a bad side. We can't ignore either.

[@Senecamarcus]

He is playing it down. Deep down he knows AI's will do unusual things. The problem a lot of people are not able to grasp and I will mention few there are a lot more: 1. Comparing current AI technology to how printing took over jobs or how cars took over horses is the worst comparison one can make, for instant recently Google DeepMind AlphaGo team worked on a chess engine and within 4hrs through ML the engine was able to beat the #1 chess engine in the world and it's not complex wins the google engine destroys one of the most powerful chess engines. And this was done in 4hrs. People do not understand how ML and AI can solve insanely complex problems and get better at it. 2. The case a lot of experts make is that these AI's are all good at repetitive tasks or tasks that are geared towards a single task exp: identifying images, learning chess, Go, running, jumping etc. But the question is what if all these features were combined into one AI? While an AI is stupid at some level and they make the argument that AI is as stupid as a new born the truth is that an AI can learn something in a second what will take a human lifetime. Even though AI's will make our life a lot easier and we will excel and prosper, we have to be honest with ourselves that this time its not the same. When cars firsr came out it didn't drive millions of kilometers on their own in simulation to learn and adopt to environment. We humans handled it. Today, AI drives millions of kilometers and goes through millions of simulations in no time to learn how to drive a car.

[@xd101dxxd8]

Kale Mussel... Exactly and if you observed his face when he was continuously being questioned you'd easily come to know that he's just trying to tackle them.

[@christinegravatt1760]

There will be programmers with super powers via piloting A.I. & other crazy nano technology. hmu #resistence

[@frankmaclow2709]

Now this bold guy is an entrepreneur which means it is in his own financial interest to tell us nothing will go wrong with AI. He needs money to be invested in his company and doesn't want to scare shareholders away. If you get in a taxi and the taxi driver tells you : "I can get you there but can't guarantee you that we won't crash on the way" you would get off his car immediately and wait for the next taxi. When the guy in front of him tells him people are concerned, he gently laughs and bisically say they watch too many sci-fi movies. It is a very suspicious behavior to say the least, considering the fact that many engineers and also other entrepreneur like Elon Musk who doesn't work on AI say it could turned out to be a disaster for humanity if we are not cautious enough.

[@ugurgoktolga7359]

What I miss in the conversation are that they did not mention about the so called "paper clip maximizer", and AI ending up in the wrong hands. Paper clip maximizer is a thought experiment which says the following: even if you don't have any malice motives when designing an AI, you can end up destroying the world. In this specific case, if you program an AI to maximize the production of paper clips, it can turn the whole world into a paper clip factory, destroying everything else. Because that was their objective function. So, even if your AI is very domain specific and bound to a repetitive task, it can be dangerous. The second point is: you think of snapchat and in the morning AI makes it up and running for you. That is nice and all. But what if you think of a very deadly weapon and AI prepares it for you within one day? What do you think will happen if such a technology ends up in North Korea? It would be nice if they talked about these issues as well.

[@DrB1900]

He's not going to say anything negative about AI, because he knows how close they are to being our overlords. :)

[@ashoknayar9527]

He's also wrong about AlphaGo's brittleness. AlphaGo was trained off of existing datasets, but AlphaGo Zero learned from no dataset, and AlphaZero can play very different games at superhuman strength so it has a form of generalized expertise that he considers there to be no progress on. Nothing to be too scared about but it's worth being part of the conversation.

[@ryanmeok9800]

that background moving when speakerput his hands up

[@woewoe2749]

Truly amazing. Loved this speaker.

[@yoram9692]

This one got some sweat on him

[@piotrkucharczyk8094]

He is sweating because he knows that GoogleAI is watching and if he stops being so optimistic about it for one second his automatic car will randomly malfunction the same day ;)

[@yoram9692]

Piotr Kucharczyk haha

[@Cghost-fh4hf]

Inspiring speech!

[@frankmaclow2709]

Cghost 3578 This guy doesn't want to scare away his investors and future shareholders. He is working on AI so he is not going to say that it could be dangerous. All the engineers and scientists across the glob, including Elon Musk, are saying AI could be the end of us if we are not careful. It's funny to see that those who say everything will be wonderful are those working on AI and who are financially linked to it. Look at the million dollar smile on this bold guy's face ! Inspiring speech ?!

[@Cghost-fh4hf]

There are two scenarios of growing AI. First - we are going to solve all our problems including aging. Second - we will be destroyed. Both are acceptable for me, because i am fed up with human stupidity. I prefer to be destroyed by smart AI than live with stupid people, to be honest.

[@stefandiestefan7204]

Great talk :)

[@4relevants]

I can see this guy as my crew member on Star Trek Enterprise

[@SuperAtheist]

What a relief! Hakuna Matata

[@PaperDragons]

So, what if 10 years from now have hundreds of thousands of these deep learning algorithms that have figured out how to do everything from writing poetry, playing games, art, and science. Then a special deep learning system is created to analyze all the existing deep learning process. From the beginning data dump to end successful end result. It could identify the common pattern or thread between the systems, identifying how and why questions are asked, understood and answered. with a goal of teaching itself to operate like a human mind.

[@MegaBaellchen]

"Now we don't have to think anymore, we just give them lots of data." 3:08 Isn't that intellectual inscest?

[@sonnguyen-bk8jw]

Not hotdog

[@Max-ey6bv]

From the way he mentioned that software engineer made so much money and also inferred AI makes coding much easier, I think he is definitely trying to get rid of normal engineers one day and only leave the smartest ones. Thus, not only labor workers will lose their jobs, the average programmer will also face challenge to keep their positions. Although the world is always going to be controlled by small parties but this direction is quite scary considering there is no law(probably never) to control the technology advancement.

[@Max-ey6bv]

He definitely also wants to get rid of those smartest engineers if he could lol. All the benefits of AI he mentioned are good for him, a rich and powerful guy who can hire whoever he wants and fire if it is necessary.

[@Max-ey6bv]

By the way I finally understand why there are so many people are craving for crowdsourcing. The cheapest way to get information from high skilled people. Students are probably the cheapest labors in the world.

[@gunillak]

Excellent

[@bas182341]

not sure which one is more scary: his sneaker game or the AI game

[@stevet9864]

So, AI learning has crossed a tipping point where it’s gone from useless to incredibly good BUT mechanization will never cross a tipping point where it can replace human function faster than we can learn new stuff. EVERYTHING is changing in AI learning, but progress and increasing comfort and wealth is business as usual.

[@Holobrine]

Both humans and computers can now learn by pattern recognition. However, computers are not yet able to translate that into rules. That will be the next step towards the singularity.

[@saf10music]

This person is probably great at what he does, smart too. He's not a philosopher, however. Neither does he have a deep understanding of people and social enviroments. He shouldn't speak on the long term issues with AI because he obviously doesn't understand them. He brings up terms like consciousness in AI, when in reality whether the AI is conscious or not makes no difference.

[@angelvarela9830]

This makes me excited to be a software engineer

[@MiahDougie1]

These guys are aliens of the sweating kind..

[@alexsmith2526]

OK now show me a bot that can do underground mining - allowing for the earth movement from day to day - or foundry work - or house painting - there are so many variables in just those three instances - each challenge is more complex and needs human skill levels - - he is referring mainly to office work - outside the box - is the way to go

[@gawayne1374]

"3:00 they get slapped, they get spanked" ... No wonder children aren't learning anything these days lol

[@MarkHidden]

"Can't even play chess" 15:09 . About that... Good joke for a few months anyways. We don't yet know the limits of this type of machine learning, testing is essential. I would like to see deepmind apply it's thinking to the cloth folding problem.

[@binhan6432]

This is making me cry. Yes, people shouldn't have the endless fear like we had for every revolution. What's wrong with people lossing jobs when their jobs can be done more economically? think about globalization - the Made in China that everyone complains, made most people able to purchase nice deisgned furnitures - IKEA, devices such as iPhones. Those Americans who lost their jobs cuz of this, had moved to other field and mostly better off. what if there was just no need for such jobs? What if people can just enjoy without work when the foods can be produced so so so cheaply??

[@ViezeKnuf]

Dem shoes.

[@MrAndrew535]

Just a brief reference to the Video title: "What "AI" Is and isn't" is of less importance than what "AI" will or will not become. At the very least, the latter question is infinitely more interesting and will ultimately determine the future of the human species. Every comment and theory on the subject so far has been poorly considered which are invariably based on anthropocentric expectations based mostly on a fairytale foundation. I have continued to point out fundamental errors in language use which serves to perpetuate this intellectual blindness in this and many other areas of science and technology. So is it more preferable for "AI" not to be conscious? Under what circumstances is that? Are there circumstances where it would be a disadvantage for humanity if AI were to develop and evolve its state of consciousness? The human frame of reference with respect to education is also erroneous in this respect and indeed self-contradictory where "AI" is measured by its ability to achieve university entry-level requirements and yet beats all qualified dermatologists in identifying innocuous looking moles as carcinomas. It is no suprise, therefore, why tradioional institutionalised "thinking" lacks the intellectual instruments necesarry to be able to predict how "AI" is likely to become manifest beyong the event horizon of the singulartity.

[@MetalHead9074]

I welcome our AI leaders.

[@xerotoninz]

Well, most of us are creative

[@ageofatheism6638]

AI will have the creative capacity millions of times that of a human eventually, so he is dreaming when it comes to humans being left with that task.We will for the most part be left to spend our time the way WE choose.I just worry that at some stage AI is the one that chooses what humans do.

[@uttaradit2]

Turn the aircon on....

[@prakadox]

Alpha Zero quite literally did what he said doesn't happen - Learnt Go, beat the daylights out of the program that learnt from millions of human games (the original alpha go) and then went on and learnt chess and beat the daylights of one of the best chess engines and shogi engines. The hyperparameters of Alpha zero are the holy grail of the world today. No one knows what will fall to deepmind's investigations this year. About creativity, hasn't he heard of the music creations of the program named Emily Howell?

[@zeromailss]

AI isn't perfect but they are already better than human and that is good enough

[@nileshbhakre7050]

I am busy while making Google AIY project with raspi

[@mhtinla]

Fancy shoes...

[@michaelnurse9089]

Do you know what they say about fancy shoes... ....fancy feet!

[@NazwaBricksZajeta]

a lil bit too fancy

[@miguelchevres4669]

Can we talk about his drip doe

[@a7i20ci7y]

I agree with pretty much everything this guy says, but there's going to be a serious problem with the economy once AI is doing all the labor and people are doing all the creativity. I can't envision an economy that can function like that, and unfortunately, we need to have a functioning economy.

[@frankmaclow2709]

This guy doesn't want to scare away his investors and future shareholders. He is working on AI so he is not going to say that it could be dangerous. All the engineers and scientists accross the glob, including Elon Musk, are saying AI could be the end of us if we are not careful. It's funny to see that those who say everything will be wonderful are those working on AI and who are financially linked to it. Look at the million dollar smile on this bold guy's face !

[@ryccoh]

It's only a distribution problem, when productivity is up immensenly society will become wealthier it doesn't matter if it's done by AI or humans we just have to rethink wealth distribution

[@a7i20ci7y]

Rykahnz therein lies the problem.

[@ryccoh]

Yeah it becomes a political problem, I hope that when people become loud enough and care enough politicians will budge. Despite the distrust in politicians this has happened pretty often historically.

[@ryccoh]

Maybe you can add some tactical strategy then

[@jeffbingaman2754]

They always sell new tech as making things cheaper. Go by a plastic part for your car and tell me how cheap it is. They're crazy expensive. totally opposite. The cheaper things are to manufacture, the more the CEOs see a bigger profit margin.

[@tokenzoken3189]

cool red shoes

[@jerm184]

SingularityNet , for new AI programmers like he describes.

[@martincasey5110]

Those kicks...

[@alexandradurrant5166]

Where do our personalities come from?

[@amadusy889]

NOTE THE SHOES

[@TheXuism]

This guy didn't get the point or try to redirect the people's worry of job loss in the future. The truth is whatever how many jobs AI create will be much less than it eliminated . The point is we need a new material wealth distribution system which is totally different with current free market and capitalism system. For example UBI ( universal basic income ) will be one of solution to solve this problem . I am not talking about politics but it is a concept reform more than a technology reform to solve the job issue.

[@MrBranboom]

Hodor peaked through the door way, there was nothing there but he stood his ground anyway. Turns out that we're all useful idiots, and the red woman just wanted to preoccupy Hodor so she could get him stabbed in the back. Don't break glass in the cabbage patch, it turns steel to hammered copper, and from what I've seen you don't want to be the one to do it. I think I saw him just a few days ago.

[@DampestCorn]

I think anyone working in the field of AI, should be required to counteract negative impacts of the technology on the society, if they can't solve the problem, no additional progress should be made. Currently it seems that AI will take a large amount of population out of their jobs (unlike industrial revolution that needed people to build/operate machines...) - not sure how we'd be able to monetize anything we do, if only the best get paid... or how you'd finance your children without their ability to get entry level jobs to pay for their stuff... Also, I entirely disagree with the creative portion - this guy lives in a bubble with some very smart people and I'm not sure he's very aware of the rest of the population... Also, our lives are full of repetitive stuff, that's how we develop skills and progress in life - if that's taken away, what else is left?

[@MrElesday]

Changing the economy and so on is not the job of the researchers, it's the job of politicians and economists

[@SMBCheeky]

Did you guys not hear of SkinVision?

[@heltok]

15:05 Can't even play chess... That statement didn't age well...

[@Marc-dx9zl]

The name Sophia I'm going to have to count it a tribute to Roberto belarmino of Tuscany who wrote, he is the one who wrote Live well and die holy. a very high-level pastors handbook. Sophia the bot. Iron and clay. We humans are the clay. The bot would be iron. Or we are the iron and the chip is the clay. Iron blood, clay silicon.

[@crueI]

He says rules but he means strategies. You can't change the rules of chess or go. Also now there's Alpha-Zero that doesn't need any game database to learn from.

[@mchapman8960]

Thrun talks about an algorithm that can recognise dog and cat faces. Can this algorithm or another algorithm distinguishes between something that looks like a cart face vs something that is a cats face?

[@DaManBearPig]

AI will help us design and engineer things beyond our wildest dreams But it’ll be terrifying to watch it Turn our society upside down

[@bene2929]

15:08 Now it can🤣

[@marziahamdard6051]

GREAT

[@elFuego29]

This one was entertaining. We have one Lex Luthor, explianing himself as a Elon Musk. Laughed

[@kartikk704]

Self driving cars won't work in New Delhi,India😂

[@fzproject6462]

*shut up and take my money!!!* GIVE ME AN AI FOR PRESIDENT ALREADY!!!

[@wol2231]

FZ Project aha Something's similar gona happen very soon. All the government work will be done by AI in near future.

[@fzproject6462]

Puffy Pony I bet it'll work way more efficient and effective, not to mention reducing corruption chance. the day machine tell us homan what to do will come soon 😂😂😂

[@mvfs]

Public ally available military technology is 60 years old when just leaked or released. Google is a public company making known advances in AI, but none of which are anywhere near AGI. Imagine the sort of computing capabilities and AI the military has secret.. TLDR; All world leaders are governed and run by AI, a simple decision making program

[@nupurnishant6911]

In case you didn't know, he has an h-index of 149!

[@galettimusic3837]

"Let them eat apps!"

[@David-mu8hn]

Plot twist: he is a robot

[@davincifyable]

I wouldn't be comfortable in a self driving car with just a camera as the sensor. No one can guarantee a 100% test accuracy when using a Neural Network(or any other Machine Learning model). I don't want to lose my life over a neural network's false prediction. I think cars with a Lidar sensor along with a camera would be much safer.

[@wol2231]

Not when all the cars runs on a decentralised system network which basically means Cars communicate with other cars at a certain specific distance. Eg: Bitcoins transactions works in a similar way :-) .

[@davincifyable]

Yeah, that would be much safer. However, I don't think the current self-driving cars communicate with other cars while making driving decisions.

[@keegsmark]

I'm game for a neural implant, next level evolution.