A true scientist! Seeing the empirical evidence in his surroundings thus realizing it's possible. These are the people we don't have enough of and truly inspire me.
It is interesting how his great-great-grandfather George Boole was ahead of his time when his work about Boolean Algebra was build on a basis of a model of how the mind works, which for him was logic. After around 150 years his direct descendant Geoffrey Hinton pursues the goal to understand how the mind works and then goes ahead of his time by revolutionizing Artificial Intelligence. The world did indeed catch up once, but he is still ahead as he keeps pushing the field with new work like Capsule Neural Networks.
@Kevvy Kim Thank you for the explanation ❤☺ But at its core, deep learning is just chained regression. Of course, errors aggregates in different layers. So, a recent paper (neural ordinary differential equations) tried to improve the "fitting" process using infinite layers, a.k.a, using equations instead of discrete layers (like in calculus, from a series of discrete regressions to a continuous measurement). It's pretty just applied math. But personally, I don't think that's how the brain works, although they call it "neural" network. Quantum biology is going on in our brains (or at least in migratory bird's brains, and in plants' photosynthesis process), and even physicist cannot fully explain anything Quantum yet. Until then, I believe we won't be able to build AGI, not without mimicking the quantum process in nature.
Well, he quit Google and warning from AI, the thing he spent his life building. If this was a sci-fi film I wouldn't have believed it, but it is happening in real life
lol, yes, but now we are only interested on AI. AI in the words of another genius is only better stats. For me standing is more important - i can sit only 2-5h a week.
Wow, this video about the Canadian genius who created modern AI has aged incredibly well! Even four years later, the impact and significance of Geoffrey Hinton's work in machine learning and neural networks are still being felt and expanded upon by researchers and developers around the world. It's amazing to think how much progress has been made in AI and deep learning since this video was published, and Hinton's contributions continue to be at the forefront of these advancements. Thank you for sharing this insightful and informative video! - ChatGPT
I'm only putting my comment here because your name is Lucas. I am a non academic but I attended a special interest group's talk at Melbourne University in the mid 80's where neural nets and parallel programing was discussed. I immediately understood the merits of this approach and thoroughly believed in it thereafter. However, as mentioned in the video, they couldn't make it work. Nevertheless from a philosophical perspective I could see clearly how it 'must' be able to work. Nowadays it is being touted as 'AI' or the means to AI. Whilst I understand what might be accomplished in the field of 'machine learning' I nevertheless seem to be in a very small minority of people who insist that 'artificial intelligence' is fundamentally impossible. Consciousness precedes intelligence. In order to build an 'intelligent' machine you must first build a 'conscious' machine. Neural nets might accomplish 'anything' but they cannot become conscious. Others think they know why machines 'can' achieve consciousness - I think I know why they absolutely can't.
@@MarkLucasProductions it is thoroughly amazing to me that anyone in 80s, let alone 60s, thought they could make "real AI".. the problem could be studied of course but the hardware simply wasn't there yet. and they knew roughly how many neurons were in a human brain. I mean there was never even a chance. did you know IBM built a 512 node supercomputer in 2001 that cost $110M that calculated at around 12 teraflops, the new xbox you can get from local supermarket this year has roughly the same calculation power. and it is still not enough for even a rat brain. the hubris of thinking they had any chance half a century ago.
@messiah yea, there's lots of issues with semantics of the words we use. people argue that intelligence requires intentionality, and that implies will or desire, i.e. feelings. although artificial neural networks can simulate emergent behavior it still boils down to programming and while we keep pushing the boundary of machine learning it never becomes AI, basically if you can explain it it's not AI. 🤷♂️
I went for a visit to Carnegie-Mellon University in 1986 and saw him work in his office. I didn't want to disturb him because he was intensely concentrated. He had a name for himself then.
Because now AI is popular topic. Its fashionable word these days. You dont know how much is the innovators in the world. Just dont get attention financial support and maybe dont want to.
Geoff is the great-great-grandson of George Boole (where "boolean" logic comes from, whose mathematical work is credited with laying the foundation of computers), quite fitting that Hinton is influential & pushing forward a field whose ancestor is fundamental to.
Sorry but he's British - born, raised and educated in Britain and that's where his career started. You can be proud of the fact that Canada helped him further his research - but you can't entirely claim him! Sorry!
With respect. Citing Dr. Hinton as a Canadian pioneer in AI is likely a fair statement but, such a statement ignores those who came before him. Donald Olding Hebb was born on July 22, 1904 in Chester, Nova Scotia where he lived until his family moved to Dartmouth when he was 16. The term AI was not in use when Dr Hebb became interested in Neural Nets but it is undeniable, his work is foundational to Dr Hinton’s. The source for where Dr. Hebb's foundational thinking was, is difficult to know unless he wrote it down somewhere. Personally my reading suggests statistics. As a human mind scans and evaluates any particular body of data and finds some truth, they have mimicked Hebb's net. Personally I feel the term AI is a misnomer. If something is intelligent, it is intelligent, not artificial! Again, with respect to Dr. Hinton real talent, Canada can be proud of; Ken Bowd Layperson Canada. PS: my logic here is the product of a 1970’s era tv show called “Connections” by James Burke. (Google “connections PBS”).
I used to attend lectures in the Carnegie Mellon Computer Science department in 1983-1985. OMG. Jeff shook up the original (symbolic) AI gurus so bad......the fear and hate was palatable. Jeff never seemed to enjoy being hated, but he behaved like he was fearless because he was convinced he was right about neural nets and statistics/math. Always with an amazingly deadpan (British?) sense of humor. The thing that impressed me the most about Jeff at that time was even the very best "traditional AI" students, post grads and facility quietly, and sometimes not so quietly, took his side and helped to fight off the fierce but unconvincing criticism leveled at him at that time. While I could understand why Jeff would flee to Toronto because of the DARPA money thing, I often wondered if he simply tired of the abuse at CMU.
Yoshua Bengio is also a Canadian Pioneer of AI. He was the supervisor for Ian Goodfellow, one of the top AI researchers in the world working at Google right now.
AI is accelerating and growing at such a remarkable pace. Who would’ve thought that two Canadian cities: Edmonton, AB and Toronto, ON, would end up being the epicentres and hubs for all things related to A.I & Machine Learning.
Not to mention Charles Thomas Bolton of U of T was the first scientist to actually found and identify a blackhole...and U of T aerospace engineers helped Apollo 13 landed back on earth safely.
character recognition was in the late 60's approximately, when postal services purchased from japan the zipcode recognizer thats already recognised the handwritten digits..
According to Jared who works at Porter Hospital in Littleton, CO they can hardwire into any ones brain to upload and download data 24/7. Jared works at Porter as a sleep specialist. These sleep studies are conducted on hotel rooms to make the experience comfortable and conducive to extracting data from patients who have sleep apnea and other sleeping disorders. If you want to take classes and graduate from a University with a PHD or doctorate you can. While you sleep your brain can be uploaded with all kinds of data and you can broaden your knowledge while you sleep. I have already designed alnd have several inventions and ideas that have been copyrighted. Look for news about these new technologies in the not to distant future.
@5:33 So, why did Geoff Hinton build a self-driving military vehicle? _FACEPALM!_ Oh wait, that's Dean Pomerleau's ALVINN project at Carnegie Mellon University in 1989. I guess that had nothing to do with whether Hinton took military money or not, after all.
@Snow 123 i see that you got the reference. UNESCO World Heritage Site where Books were burned for close to 3 months along with the Monks. Considered a Great Loss of Knowledge.
Hinton went from perceptron, to deep convolutional networks, to capsule networks, but he will end up with a network based on Graph Theory. W.T. Tutte is the little known genius - (another British transplant who found a home for research in Canada). Math grads from Waterloo know W.T. Tutte. He broke the Lorenz cipher to help win WWII. Alan Turing is a shadow compared to the towering intellect of W.T. Tutte.
It was Yann LeCunn developed convolutional neural networks but Hinton's work on deep multi-layer perceptron help laid the foundation for ConvNets to work.
Many years ago, I watched a TV programme where Geoff Hinton 'energetically ' debated John (Chinese Room) Searle over Artificial Intelligence. Very entertaining.
Hey that was University of Toronto downtown St George campus Convocation Hall he was walking by. Recently i learned Hinton been teaching AI there. No wonder my friend who was a CSC PhD student told me back in the late 1980 they were teaching neural network at University of Toronto. Hinton probably helps U ot T to be ranked 16 on Newsweek's top international universities ranking in recent decade.
Very cool, but I’d love to hear the guy’s views on how we’re going to cope with having advanced AI among us. He clearly must have given it a lot of thought!
He has a new interview with CBS released just a few weeks ago. It’s the best interview I’ve heard him give, and he discusses many of the things you’d like to hear; worth checking out!
He didn't invent A.I. he was part of its advancements. Hinton research began in 1972. Look up Shakey robot 🤖 it was the first A.I. mobile robot and it retired in 1972 when Hinton began. Check your facts people 💯
You get yourself an exoskeleton, they are developed for people who work standing so they can stay standing but put the weight onto the exoskeleton instead of their own bones. ✌️ 💙
The worst part is people like him are ridiculed first, when they need the most support. That's the price they end up paying for being smarter than others
bbbeto02 I mean sitting down decreases our life span, combined with the fact that he walks everywhere he will probably live 20 years longer than most humans
Input empty While input empty If not empty then not empty else empty : : If not empty then not empty else empty If empty then empty else empty : : If empty then empty else empty End Go-to begining
This video is interesting in two ways, first for the AI, but just as importantly for me personally is I thought my inability to sit was unique, now I see it's a common back problem! Thanks Geoffrey!
Spend time on board ship, standing is normal even when listing and during the perfect storm, standing becomes a Art form.. Dancing on waves. Thank you.
I’m waiting until they ask AI the ultimate question. How should humans live life that is truly fulfilling and the answer AI gives will be completely against our ego. Back to reality blending harmoniously with nature. I can’t wait.