Loading video...

Video Failed to Load

There was a problem loading this video. This could be due to a temporary network issue or the video might be unavailable.

Jeff Hinton dismissed gradient descent in the 1970s because he thought models would get stuck. He won the Nobel Prize in 2024 partly for being wrong about that. Every time ChatGPT gets better after a training run, it's because 1.2 billion parameters fell through a wormhole in a loss... show more

Eron

4,223 subscribers

128,060 views • 11 days ago •via X (Twitter)

Science & Technology Education News & Politics

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

Ray Didinger on Jalen Hurts' ankle injury scare late in the 3rd quarter against the Raiders: "That is the built-in worry about that, that every time he takes off, there's a chance that he could get hurt. But I still think it's something that they have to do because it just makes them a better team. There's no question about that."

Ray Didinger on Jalen Hurts' ankle injury scare late in the 3rd quarter against the Raiders: "That is the built-in worry about that, that every time he takes off, there's a chance that he could get hurt. But I still think it's something that they have to do because it just makes them a better team. There's no question about that."

SPORTSRADIO 94WIP

39,582 views • 8 months ago

Coach K says he did not retire because of the new environment in college sports. (NIL, Transfer Portal, etc...) "Instead of complaining about what it is, just accept the fact that this is a new time. And because it's not like the old doesn't mean that it's not ok."

Coach K says he did not retire because of the new environment in college sports. (NIL, Transfer Portal, etc...) "Instead of complaining about what it is, just accept the fact that this is a new time. And because it's not like the old doesn't mean that it's not ok."

Dan Patrick Show

211,876 views • 2 years ago

Gallego: He's going to destroy NATO, destroy our whole economy, all because he feels that he got slighted for not taking not getting a Nobel prize. This is a man that cares about himself, not about this country. And he will screw this whole country in order to make himself feel better in the process.

Gallego: He's going to destroy NATO, destroy our whole economy, all because he feels that he got slighted for not taking not getting a Nobel prize. This is a man that cares about himself, not about this country. And he will screw this whole country in order to make himself feel better in the process.

Acyn

200,705 views • 7 months ago

For years the smartest economists in America insisted that Warren Buffett and Charlie Munger were a statistical fluke, a one-in-a-billion accident. Then the loudest of them went to run money himself. This is Munger on how that argument actually ended: Their theory said nobody beats the market for long, that a long enough run of wins had to be luck. Berkshire kept winning anyway. So one of them, a man who had won a share of a Nobel Prize, called its record a two-sigma event, a fluke. Then a three-sigma event. Then four. "He finally got up to six sigmas. Better to add a sigma than change a theory, just because the evidence comes in differently." Rather than concede the theory was wrong, he kept making the record more impossible. Then he left the faculty and went to manage money against the very market he said no one could beat. "When this share of a Nobel Prize went into management himself, he sank like a stone." The theory kept its place in the textbooks. The man defending it didn't keep his in the market. Bookmark & listen/watch:

For years the smartest economists in America insisted that Warren Buffett and Charlie Munger were a statistical fluke, a one-in-a-billion accident. Then the loudest of them went to run money himself. This is Munger on how that argument actually ended: Their theory said nobody beats the market for long, that a long enough run of wins had to be luck. Berkshire kept winning anyway. So one of them, a man who had won a share of a Nobel Prize, called its record a two-sigma event, a fluke. Then a three-sigma event. Then four. "He finally got up to six sigmas. Better to add a sigma than change a theory, just because the evidence comes in differently." Rather than concede the theory was wrong, he kept making the record more impossible. Then he left the faculty and went to manage money against the very market he said no one could beat. "When this share of a Nobel Prize went into management himself, he sank like a stone." The theory kept its place in the textbooks. The man defending it didn't keep his in the market. Bookmark & listen/watch:

Fireside Tapes

125,269 views • 1 month ago

James Talarico Do we think the president is digging through the minutia of FCC policy because he's concerned about the Equal Time rule and how it's applied, or is he going after shows that talk about him in a way he doesn't like? Which so which one sounds more like the president? And this is the party that ran against cancel culture.

James Talarico Do we think the president is digging through the minutia of FCC policy because he's concerned about the Equal Time rule and how it's applied, or is he going after shows that talk about him in a way he doesn't like? Which so which one sounds more like the president? And this is the party that ran against cancel culture.

Blue Georgia

50,926 views • 6 months ago

One of the best hedge fund managers of the last 30 years, Joel Greenblatt, earned over 40% a year for two decades. He found a serious flaw in ordinary indexes like the S&P 500 and the Russell 1000. Because of it, the indexes automatically buy too much of the overpriced stocks and too little of the cheap ones. In this video he reveals exactly what that flaw is. "The best mutual fund of the 2000s returned 18% a year. The average investor in that same fund lost 11% a year" "Why?" "Because every time the fund started to lag, people left. Every time the market fell, people left. And every time the fund did great, people piled in right after the run." Watch how the individual investor can beat the big players ↓

One of the best hedge fund managers of the last 30 years, Joel Greenblatt, earned over 40% a year for two decades. He found a serious flaw in ordinary indexes like the S&P 500 and the Russell 1000. Because of it, the indexes automatically buy too much of the overpriced stocks and too little of the cheap ones. In this video he reveals exactly what that flaw is. "The best mutual fund of the 2000s returned 18% a year. The average investor in that same fund lost 11% a year" "Why?" "Because every time the fund started to lag, people left. Every time the market fell, people left. And every time the fund did great, people piled in right after the run." Watch how the individual investor can beat the big players ↓

Damir Akaza

1,175,222 views • 1 month ago

CEO of Citadel: "no one is more wrong than I am today", he built the most profitable hedge fund in history in this interview he explains why he hired a Russian rocket scientist, why being the smartest in the room is a mistake, and why being right 54% of the time made $90 billion Bookmark & watch it. Then read the article below - The 77-year-old formula that explains why a small edge is all you need ↓

CEO of Citadel: "no one is more wrong than I am today", he built the most profitable hedge fund in history in this interview he explains why he hired a Russian rocket scientist, why being the smartest in the room is a mistake, and why being right 54% of the time made $90 billion Bookmark & watch it. Then read the article below - The 77-year-old formula that explains why a small edge is all you need ↓

Сarm1ne

4,245,352 views • 3 months ago

Low agency in a 2 minute video: Derren Brown creates a compliance test He places 3 actors into a room The actors stand up and sit down when the bell rings He then introduces real people into the room It's fascinating AND terrifying how many people blindly copy the other people -- simply because. Mimetic low agency OS takes over. You can see from the initial look on their faces that they know it's not right -- but they copy and paste the room -- because they don't want to stand out. Some thoughts: 1. Look at the guys face at 0.50 It's the emperors new clothes moment He can see everyone is naked But he feels naked by not going along with the tribe 2. We need a word for that moment at 0.50 Is there a good english word for that moment? It's the feeling of looking wrong socially being more important than being wrong in reality. The first step to identifying that moment is having a word 3. The blind spot trap What is fascinating about this video is that nobody watches this, including myself, thinks they would do this But the odds are, that's not true... That blind spot is part of the trap. What a video.

Low agency in a 2 minute video: Derren Brown creates a compliance test He places 3 actors into a room The actors stand up and sit down when the bell rings He then introduces real people into the room It's fascinating AND terrifying how many people blindly copy the other people -- simply because. Mimetic low agency OS takes over. You can see from the initial look on their faces that they know it's not right -- but they copy and paste the room -- because they don't want to stand out. Some thoughts: 1. Look at the guys face at 0.50 It's the emperors new clothes moment He can see everyone is naked But he feels naked by not going along with the tribe 2. We need a word for that moment at 0.50 Is there a good english word for that moment? It's the feeling of looking wrong socially being more important than being wrong in reality. The first step to identifying that moment is having a word 3. The blind spot trap What is fascinating about this video is that nobody watches this, including myself, thinks they would do this But the odds are, that's not true... That blind spot is part of the trap. What a video.

George Mack

53,894 views • 1 year ago

"I think we can use AI as part of the solution if we're clever." Today Demis Hassabis was awarded the Nobel Prize in Chemistry 2024 for developing an AI model that can predicting proteins’ complex structures. He joined us earlier this year at a Nobel Prize Dialogue event and spoke about the opportunities of AI. Watch the full conversation:

"I think we can use AI as part of the solution if we're clever." Today Demis Hassabis was awarded the Nobel Prize in Chemistry 2024 for developing an AI model that can predicting proteins’ complex structures. He joined us earlier this year at a Nobel Prize Dialogue event and spoke about the opportunities of AI. Watch the full conversation:

The Nobel Prize

85,025 views • 1 year ago

AI cannot make money. ChatGPT costs billions to run. Generates almost nothing in return. So why is Amazon, Microsoft, Google and Meta tripling their AI investment every year? Because they are all investing in each other. It is a circle. Nobody is making money. Everyone is pretending they will. There is only one solution to a bubble this big. The government. Three days after Trump's inauguration — Operation Stargate. $500 billion for AI data centers. Not because AI is profitable. Because the government needs the surveillance infrastructure that AI makes possible. AI needed the government to survive. The government needed AI to control. They found each other. And you are paying for it.

AI cannot make money. ChatGPT costs billions to run. Generates almost nothing in return. So why is Amazon, Microsoft, Google and Meta tripling their AI investment every year? Because they are all investing in each other. It is a circle. Nobody is making money. Everyone is pretending they will. There is only one solution to a bubble this big. The government. Three days after Trump's inauguration — Operation Stargate. $500 billion for AI data centers. Not because AI is profitable. Because the government needs the surveillance infrastructure that AI makes possible. AI needed the government to survive. The government needed AI to control. They found each other. And you are paying for it.

Predictive History

19,635 views • 3 months ago

🚨 READ THIS TWICE. A black man gets caught jaywalking with a stash… And the judge says: " Walking while black. Case dismissed.” Not because he was innocent. Not because the evidence was wrong. But because of who he is. Let that sink in. Because the second the law stops being blind… it stops being justice. And when that happens— it won’t just affect “them.” It’ll affect YOU. This is happening WAY TOO MUCH! Let them out ...and they commit another crime. Let them out for THAT crime... they go commit another crime. Vicious cycle . Are we really okay with this? 👀

🚨 READ THIS TWICE. A black man gets caught jaywalking with a stash… And the judge says: " Walking while black. Case dismissed.” Not because he was innocent. Not because the evidence was wrong. But because of who he is. Let that sink in. Because the second the law stops being blind… it stops being justice. And when that happens— it won’t just affect “them.” It’ll affect YOU. This is happening WAY TOO MUCH! Let them out ...and they commit another crime. Let them out for THAT crime... they go commit another crime. Vicious cycle . Are we really okay with this? 👀

🎹 Ames™ 🎹

24,744 views • 5 months ago

DoorDash CEO Tony Xu says every single person at the company still does deliveries to this day, because there are 20 things that can go wrong bringing you a burrito and no way to learn them from behind a desk. "That is the DoorDash problem set, for better and for worse. I remember when we started DoorDash, I remember decomposing that there's almost like 20 mini systems, many little processes. If you imagine a checklist of just bringing you a burrito, there's 20 little things that you have to get right, because if anything goes wrong, you'd actually need to fix it." "And there's no way that you would know about this unless you actually did the deliveries yourself. This is exactly why, to this day, we still have everyone in the company dash, which is our way of saying doing deliveries. We still all do deliveries." "Because until you actually get into the physical world, until you find yourself stuck in the wrong elevator, or the wrong lobby to try to get upstairs to deliver something, until you get to the wrong alleyway for parking, until you get to the wrong place because you realize food gets made sometimes in three different stations inside a restaurant, until you actually experience those things, it's very difficult to a priori just intellectualize that and know about it."

DoorDash CEO Tony Xu says every single person at the company still does deliveries to this day, because there are 20 things that can go wrong bringing you a burrito and no way to learn them from behind a desk. "That is the DoorDash problem set, for better and for worse. I remember when we started DoorDash, I remember decomposing that there's almost like 20 mini systems, many little processes. If you imagine a checklist of just bringing you a burrito, there's 20 little things that you have to get right, because if anything goes wrong, you'd actually need to fix it." "And there's no way that you would know about this unless you actually did the deliveries yourself. This is exactly why, to this day, we still have everyone in the company dash, which is our way of saying doing deliveries. We still all do deliveries." "Because until you actually get into the physical world, until you find yourself stuck in the wrong elevator, or the wrong lobby to try to get upstairs to deliver something, until you get to the wrong alleyway for parking, until you get to the wrong place because you realize food gets made sometimes in three different stations inside a restaurant, until you actually experience those things, it's very difficult to a priori just intellectualize that and know about it."

Fireside Alpha

33,966 views • 1 month ago

Cloudflare CEO Matthew Prince says Googlebot sees 3.2x more of the web than OpenAI, and 4.8x more than Microsoft. And he worries this advantage will allow Google to run away with the AI race, with no one else being able to catch them. "For every one page that OpenAI sees, Google is seeing 3.2." "What I worry about is, because Google has this unique access to the web that nobody else has, the game might just go to them. Because at the end of the day, whoever has the most data wins in the era of AI."

Cloudflare CEO Matthew Prince says Googlebot sees 3.2x more of the web than OpenAI, and 4.8x more than Microsoft. And he worries this advantage will allow Google to run away with the AI race, with no one else being able to catch them. "For every one page that OpenAI sees, Google is seeing 3.2." "What I worry about is, because Google has this unique access to the web that nobody else has, the game might just go to them. Because at the end of the day, whoever has the most data wins in the era of AI."

TBPN

164,015 views • 7 months ago

Cloudflare CEO Matthew Prince says Googlebot sees 3.2x more of the web than OpenAI, and 4.8x more than Microsoft. And he worries this advantage will allow Google to run away with the AI race, with no one else being able to catch them. "For every one page that OpenAI sees, Google is seeing 3.2." "What I worry about is, because Google has this unique access to the web that nobody else has, the game might just go to them. Because at the end of the day, whoever has the most data wins in the era of AI."

Cloudflare CEO Matthew Prince says Googlebot sees 3.2x more of the web than OpenAI, and 4.8x more than Microsoft. And he worries this advantage will allow Google to run away with the AI race, with no one else being able to catch them. "For every one page that OpenAI sees, Google is seeing 3.2." "What I worry about is, because Google has this unique access to the web that nobody else has, the game might just go to them. Because at the end of the day, whoever has the most data wins in the era of AI."

TBPN

98,466 views • 7 months ago

Charles Hoskinson told me that he avoids talking about XRP because every time he does, he gets mobbed. This includes the debate between him and Brad Garlinghouse where Charles said "a bad crypto bill is not better than no bill. I firmly believe no bill is necessary at this juncture until we get a proper bill,"

Charles Hoskinson told me that he avoids talking about XRP because every time he does, he gets mobbed. This includes the debate between him and Brad Garlinghouse where Charles said "a bad crypto bill is not better than no bill. I firmly believe no bill is necessary at this juncture until we get a proper bill,"

Wendy O

19,168 views • 6 months ago

Here's the full clip. "I support trans women because they are my only matches on tinder. They are my future" That's a position that you end up in by choice. And that's fine. 90% of society would not date a trans woman. That means this guy is further ahead than most people. Including a lot of trans men, according to the rest of the bit. We are still the butt of the joke though, because the humor lies in the "self-depricating" way that he would be settling for a trans woman, despite that being what he obviously desires. It's particularly unsettling because of the implication it carries for those future relationships. A situation where a man feels like he is "settling" for a woman means he sees her as less worthy of respect than other, better women. That attitude is what Lundy Bancroft identifies as the attitude from which misogynistic abuse originates.

Here's the full clip. "I support trans women because they are my only matches on tinder. They are my future" That's a position that you end up in by choice. And that's fine. 90% of society would not date a trans woman. That means this guy is further ahead than most people. Including a lot of trans men, according to the rest of the bit. We are still the butt of the joke though, because the humor lies in the "self-depricating" way that he would be settling for a trans woman, despite that being what he obviously desires. It's particularly unsettling because of the implication it carries for those future relationships. A situation where a man feels like he is "settling" for a woman means he sees her as less worthy of respect than other, better women. That attitude is what Lundy Bancroft identifies as the attitude from which misogynistic abuse originates.

Stef Girl 🏳️‍⚧️

1,100,866 views • 1 month ago

Fourier wrapped a signal around a circle in 1822. Every Spotify equalizer, every noise filter in your AirPods, and every frequency layer inside ChatGPT runs on what he found. The trick is simple. Spin the signal at different speeds. When the spin matches a hidden frequency, the center of mass shifts. That shift is the answer. Claude and GPT use this same idea to separate relevant tokens from noise. 20 minutes. Bookmark & watch today. The math is older than the telephone.

Fourier wrapped a signal around a circle in 1822. Every Spotify equalizer, every noise filter in your AirPods, and every frequency layer inside ChatGPT runs on what he found. The trick is simple. Spin the signal at different speeds. When the spin matches a hidden frequency, the center of mass shifts. That shift is the answer. Claude and GPT use this same idea to separate relevant tokens from noise. 20 minutes. Bookmark & watch today. The math is older than the telephone.

Eron

57,222 views • 25 days ago

Complex numbers aren't imaginary. They're the only way to describe a wave that has both amplitude and phase at the same time - and every LLM processes language as exactly that kind of wave through its attention layers. The same math that describes quantum wave interference is what makes Fourier transforms work, which is what lets GPT process tokens in parallel instead of one at a time. Descartes named them "imaginary" before anyone understood what they were. 20 minutes. Bookmark & watch today. The confusion clears up faster than any textbook manages.

Complex numbers aren't imaginary. They're the only way to describe a wave that has both amplitude and phase at the same time - and every LLM processes language as exactly that kind of wave through its attention layers. The same math that describes quantum wave interference is what makes Fourier transforms work, which is what lets GPT process tokens in parallel instead of one at a time. Descartes named them "imaginary" before anyone understood what they were. 20 minutes. Bookmark & watch today. The confusion clears up faster than any textbook manages.

Eron

42,680 views • 9 days ago