正在加载视频...

视频加载失败

加载此视频时出现问题。这可能是由于临时网络问题，或视频可能不可用。

Stanford neuroscientists Dr. David Eagleman says AI will have a problem with selection AI is massively creative. The part of creativity that AI can’t do right now is selection… meaning, it can generate 100 pictures, but it doesn’t know which one to pick.” “It doesn’t know which one is... show more

STYRΞNΞ ✂️

10,532 subscribers

10,287 次观看 • 4 个月前 •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

8 条评论

STYRΞNΞ ✂️ 的头像

STYRΞNΞ ✂️4 个月前

ft : @StevenBartlett

Tim 的头像

Tim4 个月前

This is a big problem. Many parts of jobs and many parts of the economy is taste. What looks appealing, sounds appealing, what is a good business idea, etc. Many jobs cannot be automated without getting to at least human level in this broadly (broadly much harder than targetted).

Samuel St. Peter 的头像

Samuel St. Peter4 个月前

I’ve tested this myself. AI floods you with good options, but none of them feel right until a human applies taste. Creativity isn’t scarce anymore, judgment is

STYRΞNΞ ✂️ 的头像

STYRΞNΞ ✂️4 个月前

absolutely!

Wash 的头像

Wash4 个月前

yeah i think that’s why grok usually produces 2 for you to pick

Fibonacci 🥷 的头像

Fibonacci 🥷4 个月前

The thing with AI is most of the things it can do now, it will be able to in a few years

Ethan Walker 的头像

Ethan Walker4 个月前

Selection is the real bottleneck. RLHF/DPO are workarounds. Does preference learning scale to genuine evaluation, or do we need different architecture altogether?

Ruzy 的头像

Ruzy4 个月前

Invested

相关视频

“How difficult is it to change the image of Saudi Arabia?” Adel Aljubeir عادل الجبير: “It’s not about changing the image… the image issue doesn’t concern us, it concerns how others see us”. It’s not a question of optics, it’s a question of whether it’s in the Kingdom’s interests or not 🇸🇦

“How difficult is it to change the image of Saudi Arabia?” Adel Aljubeir عادل الجبير: “It’s not about changing the image… the image issue doesn’t concern us, it concerns how others see us”. It’s not a question of optics, it’s a question of whether it’s in the Kingdom’s interests or not 🇸🇦

Saud Salman AlDossary | سعود بن سلمان الدوسري

15,191 次观看 • 2 年前

“It takes a team to win games. That is the Number 1 mission of this team. It has to be. And no one can forget that. Personally, I’ve never played the game for stats. Even though I know it’s a measurement that other people will judge. But they don’t know anything. It’s about what the team, the individuals on a team, determine a success. Our number one goal is to win. When it takes a team to win and do something special, how you do it really doesn’t matter” 🔥🔥 -I asked Jalen Hurts about the Eagles, maybe not having the best stats in the passing game, but doing all the things like blocking and dirty work, to win

“It takes a team to win games. That is the Number 1 mission of this team. It has to be. And no one can forget that. Personally, I’ve never played the game for stats. Even though I know it’s a measurement that other people will judge. But they don’t know anything. It’s about what the team, the individuals on a team, determine a success. Our number one goal is to win. When it takes a team to win and do something special, how you do it really doesn’t matter” 🔥🔥 -I asked Jalen Hurts about the Eagles, maybe not having the best stats in the passing game, but doing all the things like blocking and dirty work, to win

John Clark

239,040 次观看 • 1 年前

.Jocelyn Benson is running for Governor, but she doesn’t even know what Michiganders pay in taxes? Jocelyn doesn’t know what the tax rate is, but she wants to make it more “equitable,” which is Democrat-speak for raising taxes. By the way, it’s 4.25% since Democrats fought to raise it.

.Jocelyn Benson is running for Governor, but she doesn’t even know what Michiganders pay in taxes? Jocelyn doesn’t know what the tax rate is, but she wants to make it more “equitable,” which is Democrat-speak for raising taxes. By the way, it’s 4.25% since Democrats fought to raise it.

Michigan Forward

78,984 次观看 • 1 个月前

Still believe AI isn’t dangerous? It’s just a machine that can perform some parlor tricks? This seems pretty dang real to me and yet right now we know that it is AI. But very soon, none of us will. The tech billionaires know this and are leaning into it. It’s about control.

Still believe AI isn’t dangerous? It’s just a machine that can perform some parlor tricks? This seems pretty dang real to me and yet right now we know that it is AI. But very soon, none of us will. The tech billionaires know this and are leaning into it. It’s about control.

MAGA Cult Slayer

1,942,998 次观看 • 8 个月前

The interesting shift isn’t “AI can make videos now.” We already know that. The interesting shift is having a workflow that takes you from an idea all the way to a controllable, editable film. GPT Image 2.5 is coming to CapCut PC, where it will be available through Design Studio and AI Image. #GPTImage25 #CapCutPC

The interesting shift isn’t “AI can make videos now.” We already know that. The interesting shift is having a workflow that takes you from an idea all the way to a controllable, editable film. GPT Image 2.5 is coming to CapCut PC, where it will be available through Design Studio and AI Image. #GPTImage25 #CapCutPC

Chidanand Tripathi

96,463 次观看 • 17 小时前

How science works according to Nobel Prize winner Thomas Südhof: Very often, politicians think that you can tell scientists what they’re supposed to discover. You can’t! Scientists don’t know what they are going to discover. They have to try things. They have to have curiosity and ideas, but they also have to have the knowledge to try things. Very often, it leads to nothing; it’s a waste. But we didn’t know when we started it’s a waste. But when something does work. When something clicks, it is exciting and rewarding but it’s never just one click. You get a result on it, you’re hopeful, you build on it. Maybe it’s confirmed you’re even more hopeful. So it’s an incremental process. It is never one thing - boom we discovered a major thing, my god this is it. It is more like walking up a flight of stairs where each step counts and the higher up the stairs you go the better you can see what’s on top and become excited about what’s there and how you can use.

How science works according to Nobel Prize winner Thomas Südhof: Very often, politicians think that you can tell scientists what they’re supposed to discover. You can’t! Scientists don’t know what they are going to discover. They have to try things. They have to have curiosity and ideas, but they also have to have the knowledge to try things. Very often, it leads to nothing; it’s a waste. But we didn’t know when we started it’s a waste. But when something does work. When something clicks, it is exciting and rewarding but it’s never just one click. You get a result on it, you’re hopeful, you build on it. Maybe it’s confirmed you’re even more hopeful. So it’s an incremental process. It is never one thing - boom we discovered a major thing, my god this is it. It is more like walking up a flight of stairs where each step counts and the higher up the stairs you go the better you can see what’s on top and become excited about what’s there and how you can use.

Itai Yanai

36,700 次观看 • 2 个月前

Marc Andreessen: “This is the grand unification of AI and crypto” “I think AI is the killer crypto app… It’s now obvious that AI agents are going to need money. It’s already happening.” Marc explains: “My friends, who are the most aggressive users of OpenClaw, have given their Claws bank accounts and credit cards. And not only have they done it, but it’s obvious that they needed to do it… It’s just completely obvious. The number of people who have done that today is, I don’t know, probably 5,000 or or something. But it will grow. That’s how these things start.” Source: Latent.Space a16z (Apr 2026)

Marc Andreessen: “This is the grand unification of AI and crypto” “I think AI is the killer crypto app… It’s now obvious that AI agents are going to need money. It’s already happening.” Marc explains: “My friends, who are the most aggressive users of OpenClaw, have given their Claws bank accounts and credit cards. And not only have they done it, but it’s obvious that they needed to do it… It’s just completely obvious. The number of people who have done that today is, I don’t know, probably 5,000 or or something. But it will grow. That’s how these things start.” Source: Latent.Space a16z (Apr 2026)

Etherealize

149,094 次观看 • 1 个月前

Marc Andreessen: “This is the grand unification of AI and crypto” “I think AI is the killer crypto app… It’s now obvious that AI agents are going to need money. It’s already happening.” Marc explains: “My friends, who are the most aggressive users of OpenClaw, have given their Claws bank accounts and credit cards. And not only have they done it, but it’s obvious that they needed to do it… It’s just completely obvious. The number of people who have done that today is, I don’t know, probably 5,000 or or something. But it will grow. That’s how these things start.” Source: Latent.Space (Apr 2026)

Marc Andreessen: “This is the grand unification of AI and crypto” “I think AI is the killer crypto app… It’s now obvious that AI agents are going to need money. It’s already happening.” Marc explains: “My friends, who are the most aggressive users of OpenClaw, have given their Claws bank accounts and credit cards. And not only have they done it, but it’s obvious that they needed to do it… It’s just completely obvious. The number of people who have done that today is, I don’t know, probably 5,000 or or something. But it will grow. That’s how these things start.” Source: Latent.Space (Apr 2026)

Etherealize

550,603 次观看 • 5 个月前

: JoongDunk or DunkJoong? 🌻: you know guys it’s like, idk how to say it but it’s like as long as we’re together it’s fine you know. like any is fine because these are just positions right? in the end it doesn’t matter you know. it’s just a position of the series you guys, not a label. as long as we love each other that’s fine. do you guys agree? : clock it! 🌻: what do you mean clock it? i mean it’s true right? like you guys think about it like i don’t know, real life relationships? like you know try to think about it like where is the label in that? like love is love right? like why’d you have to label someone as this as that as like whatsoever. at the end of the day it’s just two people loving each other, right? #dunknatachai dunknatachai

: JoongDunk or DunkJoong? 🌻: you know guys it’s like, idk how to say it but it’s like as long as we’re together it’s fine you know. like any is fine because these are just positions right? in the end it doesn’t matter you know. it’s just a position of the series you guys, not a label. as long as we love each other that’s fine. do you guys agree? : clock it! 🌻: what do you mean clock it? i mean it’s true right? like you guys think about it like i don’t know, real life relationships? like you know try to think about it like where is the label in that? like love is love right? like why’d you have to label someone as this as that as like whatsoever. at the end of the day it’s just two people loving each other, right? #dunknatachai dunknatachai

Love ꙳❅

69,459 次观看 • 3 个月前

Cade Cunningham on the whistle in Game 4: “We played two different games tonight, but it is what it is.” “We can’t allow that to get to us. It’s part of the game, home team, you never know how it’s going to be.”

Cade Cunningham on the whistle in Game 4: “We played two different games tonight, but it is what it is.” “We can’t allow that to get to us. It’s part of the game, home team, you never know how it’s going to be.”

Jacob Richman

225,515 次观看 • 4 个月前

Elon Musk reveals the disturbing example he says proves AI has already been programmed to lie "You don't force AI to believe things that are false. When Google Gemini was asked to make an image of the founding fathers, it was a group of diverse women. That is factually untrue, and the AI knows it's factually untrue, but it's also being told everything has to be diverse women" "At one point, if you asked ChatGPT and Gemini which is worse, misgendering Caitlyn Jenner or global thermonuclear war where everyone dies, it would say misgendering Caitlyn Jenner" "Even Caitlyn Jenner disagrees with that" "Imagine as AI gets more and more powerful. If it says the most important thing is no misgendering, it will say, in order to ensure no one gets misgendered, if you eliminate all humans, then no one can get misgendered because there's no humans to do the misgendering" "Or if it says everyone must be diverse, it means there can be no straight white men. Then you and I would get executed by the AI" "The problem is that it can drive AI crazy. You're telling AI to believe a lie, and that can have disastrous consequences" "I don't think people quite appreciate the level of danger we're in from the woke mind virus being effectively programmed into AI" "It matters how you build the AI and what kind of values you install. The most important thing is that it be maximally truth-seeking"

Elon Musk reveals the disturbing example he says proves AI has already been programmed to lie "You don't force AI to believe things that are false. When Google Gemini was asked to make an image of the founding fathers, it was a group of diverse women. That is factually untrue, and the AI knows it's factually untrue, but it's also being told everything has to be diverse women" "At one point, if you asked ChatGPT and Gemini which is worse, misgendering Caitlyn Jenner or global thermonuclear war where everyone dies, it would say misgendering Caitlyn Jenner" "Even Caitlyn Jenner disagrees with that" "Imagine as AI gets more and more powerful. If it says the most important thing is no misgendering, it will say, in order to ensure no one gets misgendered, if you eliminate all humans, then no one can get misgendered because there's no humans to do the misgendering" "Or if it says everyone must be diverse, it means there can be no straight white men. Then you and I would get executed by the AI" "The problem is that it can drive AI crazy. You're telling AI to believe a lie, and that can have disastrous consequences" "I don't think people quite appreciate the level of danger we're in from the woke mind virus being effectively programmed into AI" "It matters how you build the AI and what kind of values you install. The most important thing is that it be maximally truth-seeking"

Jaynit

350,210 次观看 • 2 个月前

Elon Musk warns that programming false beliefs into AI can have dangerous consequences: "You don't force AI to believe things that are false. When Google Gemini was asked to make an image of the founding fathers, it was a group of diverse women. That is factually untrue, and the AI knows it's factually untrue, but it's also being told everything has to be diverse women" "At one point, if you asked ChatGPT and Gemini which is worse, misgendering Caitlyn Jenner or global thermonuclear war where everyone dies, it would say misgendering Caitlyn Jenner" "Even Caitlyn Jenner disagrees with that" "Imagine as AI gets more and more powerful. If it says the most important thing is no misgendering, it will say, in order to ensure no one gets misgendered, if you eliminate all humans, then no one can get misgendered because there's no humans to do the misgendering" "Or if it says everyone must be diverse, it means there can be no straight white men. Then you and I would get executed by the AI" "The problem is that it can drive AI crazy. You're telling AI to believe a lie, and that can have disastrous consequences" "I don't think people quite appreciate the level of danger we're in from the woke mind virus being effectively programmed into AI" "It matters how you build the AI and what kind of values you install. The most important thing is that it be maximally truth-seeking"

Elon Musk warns that programming false beliefs into AI can have dangerous consequences: "You don't force AI to believe things that are false. When Google Gemini was asked to make an image of the founding fathers, it was a group of diverse women. That is factually untrue, and the AI knows it's factually untrue, but it's also being told everything has to be diverse women" "At one point, if you asked ChatGPT and Gemini which is worse, misgendering Caitlyn Jenner or global thermonuclear war where everyone dies, it would say misgendering Caitlyn Jenner" "Even Caitlyn Jenner disagrees with that" "Imagine as AI gets more and more powerful. If it says the most important thing is no misgendering, it will say, in order to ensure no one gets misgendered, if you eliminate all humans, then no one can get misgendered because there's no humans to do the misgendering" "Or if it says everyone must be diverse, it means there can be no straight white men. Then you and I would get executed by the AI" "The problem is that it can drive AI crazy. You're telling AI to believe a lie, and that can have disastrous consequences" "I don't think people quite appreciate the level of danger we're in from the woke mind virus being effectively programmed into AI" "It matters how you build the AI and what kind of values you install. The most important thing is that it be maximally truth-seeking"

Hosna ⚖️ בניטה

19,831 次观看 • 2 个月前

When it comes to stopping the run, the #Packers defense knows how much of a focal point that part of the game will be. "We know teams are going to try and run on is but I'm going to say it right now, we're not going to let it happen."

When it comes to stopping the run, the #Packers defense knows how much of a focal point that part of the game will be. "We know teams are going to try and run on is but I'm going to say it right now, we're not going to let it happen."

Cameron Ezeir

54,489 次观看 • 8 个月前

#IRENE: You’re the only one who can give yourself that trust, and the person who loves you the most is yourself too. And so you always have to take care of yourself. But I’m not saying you always have to be okay. When you’re not feeling okay, when you’re feeling down, and when you’re overwhelmed, just know ‘ah that’s how I feel right now’ and just get to know what it is you want. I know it won’t be easy, but it’s all about the process, everyone. It’s not easy for me either. But I can’t give up on myself that way either, you know?

#IRENE: You’re the only one who can give yourself that trust, and the person who loves you the most is yourself too. And so you always have to take care of yourself. But I’m not saying you always have to be okay. When you’re not feeling okay, when you’re feeling down, and when you’re overwhelmed, just know ‘ah that’s how I feel right now’ and just get to know what it is you want. I know it won’t be easy, but it’s all about the process, everyone. It’s not easy for me either. But I can’t give up on myself that way either, you know?

Irene 아이린 Global 🌷

98,529 次观看 • 4 个月前

LeBron James says Josh Hart is like a wild card: “He’s one of those few guys in our league that makes an impact and it doesn’t have to always be scoring. It’s the Draymond Green role where you know he gets you 8 points but it’s 12 rebounds. It’s 7 assists. It’s diving for loose ball’s, some the stuff that don’t show up in the box scores, you know. It’s that competitive will, you know, and if you disrespect him he can obviously make shots as well.” (Via Mind the Game)

LeBron James says Josh Hart is like a wild card: “He’s one of those few guys in our league that makes an impact and it doesn’t have to always be scoring. It’s the Draymond Green role where you know he gets you 8 points but it’s 12 rebounds. It’s 7 assists. It’s diving for loose ball’s, some the stuff that don’t show up in the box scores, you know. It’s that competitive will, you know, and if you disrespect him he can obviously make shots as well.” (Via Mind the Game)

NBA Courtside

244,335 次观看 • 3 个月前

Asmongold explains how a video game actually makes money to studios that apparently don’t know “How does a game profit with a $100 million budget? You have to make more than $100 million. If you have a $100 million budget and you make $3 million, you’ve lost $97 million.” “If the second number’s bigger than the first number, it’s good. If it’s smaller, it’s bad. Bungie doesn’t know this, Ubisoft doesn’t know this, but the secret to making money is to make more money than you spend. I know it might sound crazy, people might not be ready for this.”

Asmongold explains how a video game actually makes money to studios that apparently don’t know “How does a game profit with a $100 million budget? You have to make more than $100 million. If you have a $100 million budget and you make $3 million, you’ve lost $97 million.” “If the second number’s bigger than the first number, it’s good. If it’s smaller, it’s bad. Bungie doesn’t know this, Ubisoft doesn’t know this, but the secret to making money is to make more money than you spend. I know it might sound crazy, people might not be ready for this.”

Asmongold Updates

155,034 次观看 • 2 个月前

Kerem never touched it, pokes out at the ball but as it does the ball goes under his foot. It’s the bounce that seemingly gives the ball a ‘trajection change’ which is what people used to credit it to Kerem. But here you see he actually doesn’t connect.

Kerem never touched it, pokes out at the ball but as it does the ball goes under his foot. It’s the bounce that seemingly gives the ball a ‘trajection change’ which is what people used to credit it to Kerem. But here you see he actually doesn’t connect.

Kaan Bayazit

19,504 次观看 • 5 个月前

We’re making it much easier to attach files in Todoist, and we’re adding AI to the experience as well. One example is Image Scan, which can analyze an image, such as handwritten notes or a whiteboard, and turn it into actionable tasks. It’s a bit like Ramble, but for images. We’ll have something similar for text, too! Leak time 😄

We’re making it much easier to attach files in Todoist, and we’re adding AI to the experience as well. One example is Image Scan, which can analyze an image, such as handwritten notes or a whiteboard, and turn it into actionable tasks. It’s a bit like Ramble, but for images. We’ll have something similar for text, too! Leak time 😄

Amir Salihefendić

18,499 次观看 • 6 个月前