Видео

One thing is clear, we need voices beyond those on the spectrum or those posed to get wealthy making decisions about AI's place in society.

I wonder why Atul Arya keeps swinging from left to right with his chair, like a teenager. Looks unprofessional. Also, we don't see Prof. Vaclav Smil do something similar.

We have got to put liberal arts and humanities back into the standard education curriculum. We are witnessing the unintended impact of 3 decades pushing STEM. None of these people have any capacity to consider whether we should be developing AI. They can only think in terms of it going rogue and killing us. That is NOT the concern. The damage to society and the individual is not coming from the AI, it is coming from how the AI is changing us, changing society, and changing work. They have a one dimensional view of intelligence as the memorization and proper production of data. This is not intelligence. The entire computer science industry has been put on a pedestal as geniuses. They are not. They are subject matter experts who have FAR too much power over society. They are not bastions of wisdom as evidenced by the fact that they cannot think in abstract, philosophical, or emotional terms and these are all equally important aspects of intelligence.

Patrick, thank you for being you. Something I've noticed among all Podcasters, professionals, scientists, specialists ect ect ect. Is that none of you ever speak to the poor. Not all poor people are stupid.

He needs to study bitcoin

I love to hear common sense logic.......especially in this perpetual state of confusion.

At 1:18:46 LeCun claims that ChatGPT was no suprise as the basic technology had existed for years. But back in Jan 2022 (less than 1 year before the release of ChatGPT) LeCun publicly claimed that even GPT-5000 would not be able to understand that a phone sitting on a table would move as the table was pushed. I think he really lacks the humiliy necessary to learn from failed predictions to get better at it. ruclips.net/video/sWF6SKfjtoU/видео.html

This was a pointless debate because the phrasing of the resolution was far too loose. The resolution can be synthesized as "is there a an existential threat from AI?" Nothing about time line or when it might happen or degrees of threat. Everyone on the panel stated that there was an existential threat no matter how small. That means if the audience was paying attention they had to vote for the resolution. Only people in the audience who believed there was a definite 100% chance of there being no threat could vote against the resolution. And as no one in there right mind can absolutely claim that, then they could not vote against the resolution. The only way the resolution can be countered is to cease to develop AI further before it becomes AGI or beyond. And as all debaters said there is no good reason to do this right now, then just like all other human endeavours we will continue to innovate ourselves into oblivion. Go humanity - rah, rah rah!

Linfa Wang is China born Chinese correct?

Agree on many topics with him (although I'm far more optimistic than he is) but the fact he doesn't have a phone / email and that he said that really no innovation happened over the last 50 years tells me how blatently out of touch he is with the world of tech. AI had a renaissance that "started" back in 2012 (and even bigger one started in 2017 with transformers) and we've had so much innovation happen just in the field of AI (e.g. diffusion models are a recent innovation).

You mean that bat virus that somehow has a human furin cleavage site, that only lives in bats over a thousand miles away, but also lives in the bat coronavirus lab down the street from the market, that has a history of accidents and being reported as unsafe, which was denied a US grant for bat coronavirus gain-of-function research with humanized mice in 2018? Right, it's only circumstantial evidence. The non-existent animal carrier is clearly the most likely explanation. I mean, it's not like the authorities have any motive to cover it up. Hmm, how much G-o-F research money is on the line?

Brought to you by the CCP

First of all, very good debate! Thank you for sharing! This gave very good information. All agree that the Huanan market is a central place for the spread and is either a progenitor place (start of the spread in humans) or an amplification place (a “super spreader” event). This result in two different paths for the virus to have spread. 1. A person at the Wuhan Institute of Virology was infected and either directly or indirectly infected one or more people at the Hunan Market. The people then infected a number of animals at the market. 2. An animal at the market infected other animals and one or more animals infected one or more humans, which then infected other humans. Both are possible, we don’t have enough data to definitively rule,out one or the other. But there is a difference in the likeliness. There have been numerous zoonotic events historically, including SARS 1, Ebola, HIV etc. There have been lab leaks before, but very rarely of dangerous viruses and never of a completely new dangerous virus. (With the possible exception of Marburg, depending on if you consider that a lab leak or a zoonotic infection). This is important since, as Alina stated, it informs us where we should focus our resources to prevent future events. If we take this one step further, it’s comparatively easy to “plug” a lab leak. It very hard, almost impossible, to prevent a zoonotic event. It seems like Alina in particular is proposing harder regulation of virus research. Although I consider it important to have adequate safety protocols in place, a harder regulation will likely have the exact opposite effect of what we want. It will make it harder and more costly to research viruses and prepare adequately for the next zoonotic event and the next pandemic, which will happen regardless of tight controls of lab work. History proves that beyond any doubt.

GPT-4 is one of the most advanced AI systems in the world, and it supposedly is extremely intelligent. I asked GPT-4, "do you understand the meaning of the words you generate ?", and GPT-4 said, "no I don't understand the meaning of the words I generate, my responses are based on statistical correlation, not understanding".

Bill gates sir❤❤❤❤❤❤❤❤❤❤❤❤❤❤❤❤❤❤❤❤❤❤❤❤❤❤❤❤❤❤❤❤❤❤❤❤❤❤❤

No debate is necessary. It was a natural spillover event several miles away from the lab. The Chinese authorities muddied the waters and unfortunately here we are again with some people -- especially that idiotic NY Times opinion piece that got everything wrong -- continuing to insist on a lab leak that didn't happen.

Someone stop this misinformation please

In Australia they are damaging people's properties with their method of mining gas. Their properties become unless.

lol, teck fail for pro/con vote at the end , for sure ai will be ... free of glitches, hahahaha

TWiV Special: How the pandemic began in Nature, in 5 key points ruclips.net/video/rZ1FGCPenns/видео.html

Alina Chan is a whistleblower. A very competent scientist ... nevertheless a personality with conspiratorial leanings.

Great debate! Thank you, subscribed! I'm on the con side but probability(AI existential impact on humanity)< prob(meteorite) seems low. I'd guesstimate more like the risk of getting into a car crash (I/we can also influence it by driving safer). I won't stop driving because of the risk of an accident though it's substantial. And really I may not have a choice because driving solves many of my other problems. Actually more worried about the carbon footprint than the accident risk. As I am for AI.

It's energy reduction that matters first, refrigeration biggest one by far 3900 twh of electricity. Airjoule is the solution to reducing cooling energy loads by 75% and can eliminate refrigerants while making fresh water. These units are in production, company just went public, big boys like ge bought in heavy, they're running with the biggest players in the game catl, Bosch, and carrier paid huge to be first to manufacture using their tech. This isn't hype, it's going to be everywhere within the decade because if you can make an AC unit, fridge, heat pump etc, be 75% more efficient than everyone else and make clean water... Btw translate the savings airjoule would do to our energy grid and that equates to a reductio of over 1 billion tonnes of carbon. Hoping like-minded people find this and like me want to invest in these companies not just for profit but for humanities success. AIRJ / AIRJW for those interested

Tegmark gets a chance for questions of the opposition and asks like 59 of them.

You need to make a much more cogent argument as to from where and why a self-preservation goal will enter the AGI equation. This is the THE pivotal risk factor.

Reasoning a bit off. Subservience is not an emotion and doesn’t require emotions. It’s a behavioral pattern which simply requires following all instructions.

Here’s my summary: while the lab leak hypothesis is entirely possible it’s not likely, and that based on the evidence we have now, the zoonotic origin hypothesis is stronger and more likely. While this could change with new evidence, so far it has not. This is the most reasonable, nonpartisan summation anyone can make after listening to these brilliant researchers.

Wow , is it a coincidence that the same percentage of voters who believe that the KKK was just misunderstood, that the earth is flat and that smoking tobacco does not cause cancer is around 37%? I think not!

I love Munk debates and was really excited about this one but... the arguments are - pardon me - lacking of substance and somewhat lazily construed. Disappointed 😞 I think Jerry Seinfeld made a clearer statement against AI in his commencement speech

On October 14, 2016 I attended a lecture by Vaclav Smil in Vancouver BC. Professor Smil does not live on a faculty salary, I believe he derives a substantial proportion of his income from his publication royalties and thus relies heavily on frequent public appearances to promote his book sales. I was amazed at how bad his English was - for a man who fled Czechoslovakia in 1967 and had ostensibly been living almost 50 years in North America. I was disappointed by his failure to grasp the fundamentals of good public speaking to a general audience: 1. he spoke much too rapidly, 2. whenever he presented a slide he continued to yammer instead of pausing for a few moments of silence to allow our collective visual-cortices to "digest" the contents of the slide, 3. he failed to anticipate and deliver rebuttals to theories that ran contrary to his own. The crux of Smil's thesis, presented that evening, seemed to be that virtually all of modern scientific discovery occurred during just two generations (between 1880 and 1914) and the so-called "progress" during the past 100 years has been merely innovative refinement of those fundamental discoveries. In other words, the discoveries of transistors, radar, television, electron microscopes, superconductivity, atomic fission and fusion, the structure of DNA, the blue l.e.d., etc. are of such minor importance as to be scarcely worth mentioning. After the presentation - during his departure from the hall - I listened to Smil field a few questions from members of the audience; I found his arrogance stifling and the insulting way he dismissed some of the questions to be somewhat distressing.

Yeah, I'm gonna trust a govt fund baby (9 govt grants) who produces poor studies for Global Think Tanks whose purpose is to bypass the will of the people through technocratic authoritarianism (epidemiological models aren't factual - they're statistics [there are lies, damned lies and statistics] and never prove causation, they find signals that require hypothesis and clinical testing.) I've studied epidemiology, from its origins with John Snow to the most famous epidemiologists like Bradford Hill and understand that the manner with which epidemiology is deployed is fraudulent. It can never prove or disprove causation. BTW, there's no such thing as a one size fits all drug that can be universally given to an entire population - a portion will always be maimed or killed due to the genetic diversity of people. As every drug has a NNT and NNH, these Injectable drugs are no different, they have a NNV: the number of people needed to be injected before one person benefits. Pfizer is 142 and Moderna 88, which means the vast majority people will receive no benefit from these drugs. I'm always shocked how propagandized the average "academic" is and how little they understand about immunology, infection and biology. A freakin economist who builds models based on numerous fallacies and false assumptions is hysterical. I've studied infectious disease for over 10 years and can absolutely say that studies like this are garbage. Tuberculosis has a higher death rate than c19 and the vast majority of people are naturally immune to TB. Only 5-10% of susceptible people have a life long possibility of developing TB. Every infection works this way without exception. Majority are immune and only subsets of groups carry a risk of infection and an even smaller group, risk of death. The host pathogen paradigm elucidates the true nature of infection and real risks associated with it, placing the host, not the pathogen, at the center of the equation and the determiner of the outcome.

Sorry, it seems the arguments from Yan and Melanie was completely emotional and ridiculous (especially Melanie arguments were disaster, it seems she did not even get the idea of from where the power of neural networks is coming)

Why is there no mention of nuclear?

right

BENEVOLENT AI is mandatory, imho

Trump poses an existential threat to our democracy, Putin poses an existential threat to mankind, Musk is trying to climb into top three and maybe there.

March 2024. This is just the tip of the AI iceberg. Wikimedia gone feral. Putting AI in charge of anything, or to accept its advice is bizarre - THAT is the problem. I was interested to be better informed, and I still am - but I am now biased against the use of AI. It is essentially PROGRAMMED by HUMANS and recently in 2024 these particular humans programmed an AI system called GERONIMO, to remove white people from the internet world. AI was used to remove pictures of white people and changed the requests of people who wanted real information. Censoring by sleight of hand is extremely dangerous. This was a deliberate attempt at the worst form of racism. This borders on a genocidal wish. Is this anti-white attack by Geronimo, acceptable? The programmers were so incompetent, thankfully they revealed what they were attempting to do and let the cat out of the bag. He assures us that AI will be subservient - why aren't we teaching history? Just last century we have seen well documented records, films, and books about genocidal maniacs with their SUBSERVIENT minions trying to eliminate ALL Jewish people, and others, as a rule of law, and sadly we see certain Arabs STILL attempting this as they did with Hitler in the 1940s, yet this real history is being denied and revised. Who is going to feed our history to the AI? Is this what we want? Regards.

Yann LeCun might be the dumbest smart person I've ever seen

I’m amazed that Yoshua Benggio is supposed to be an AI expert, but he doesn’t seem to understand that all AI systems, no matter how advanced they are, at their core they are nothing more than a bunch of mathematical formulas and equations, that enable computers to execute tasks. Since the field of AI was founded in 1956, that’s how it worked. You “plug” digital data into these equations, and you get an output. This is one such equation: W:= W-a θJ/θW. Which means that Yoshua Benggio believes that this equation, and equations like this, will become sentient and wipe out humanity. This is barking mad

Too bad sam harris wasnt in this debate.

Sounded like melanie was complaining and griping.😂

Why are all of these guys foreigners

AI is going to decide what is information and what is misinformation. Great. That makes me feel so much better.

Mostly side with the pro case, but Bengio was so annoying how many times he kept interjecting, he never let anyone finish. It's clear there was a log of pent up frustration on both sides. Not much understanding gained or progress made, just a lot of venting.

It's pretty cool to see high-profile technical professionals debate each other; this is also why I'm looking forward to seeing LazAI soon, as they are ready to take the next step in the field of AI matching with decentralization, and they're also one of the participants in the AI battle that will be held at Ethereum Denver, hosted by MetisFest.

we should teach the debators the A B C of how to use slides and deliver speeches! I would expect people at this level either place on photo on the screen or not at all. If anyone working for me present this way, I'll send them to take a communications class.

Who invited Karen?

After listening to these people... I am totally worried

And they didn't note that these systems are trained on Human information amd knowledge and these A.Is can just amplify our biases or imperfection

When LeCunn says "the good guys will just have better AI than the bad guys" I can't help but think about why he is assuming the world thinks of Meta and OpenAI and Google as the good guy :| I'm much more worried about mega corps ruining the world with AI than about terrorists honestly.