Video yükleniyor...

Video Yüklenemedi

Bu video yüklenirken bir sorun oluştu. Bu geçici bir ağ sorunundan kaynaklanıyor olabilir veya video kullanılamıyor olabilir.

Ana Sayfaya Dön

Vals AI CEO Rayan Krishnan on why sovereign AI makes Cold War-style verification necessary: "From my very idealistic perspective, I'm surprised to see so much investment in sovereign AI." "You could probably consolidate a lot of these efforts. But it seems like that's not the world we're in or... the one we're headed towards. There are actually increased efforts to build AI in a sovereign way." "So that takes having a shared language to communicate about what the framework for evaluations is, and where we're going to collectively align around the risks." "There's a lot to learn from nuclear here. Reagan had this line, 'Trust but verify.' We're starting to see signs of trust in that Xi Jinping and Trump are going to be meeting... But there is no clear way to actually do the verification part of this." "Having the shared language of evals will allow us to say things like, you have the right number of nuclear warheads. In that example, there were also flyovers, a mechanism by which a country could audit another country's nuclear stockpile." "If there's concern about the societal or even existential risk of AI, it will necessitate us constructing this shared language of evaluations to do the verification process." Rayan Krishnan Jennifer Lishow more

a16z

1,075,411 subscribers

34,808 görüntüleme • 2 gün önce •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

8 Yorum

Shinra profil fotoğrafı

Shinra2 gün önce

comparing ai verification to nuclear weapons inspections sounds smart until you realize you can't actually fly over and inspect an AI model. with nukes you count warheads and verify storage. with ai you can run a benchmark and the company just... shows you whatever they want you to see. shared evaluation language is fine but it doesn't solve the fundamental problem: you can't verify a black box model's true capabilities without the model owner letting you see them. this is just "please trust us we're aligned" dressed up as geopolitical necessity

Don · fireply.ai profil fotoğrafı

Don · fireply.ai2 gün önce

trading side has the same issue, everyone wants shared benchmarks until their own model is the one getting graded down

tinku profil fotoğrafı

tinku2 gün önce

One lesson I kept was to make the extra uncertainty visible. Writing the trade-off behind an earlier entry helped.

AI Mastery Guide profil fotoğrafı

AI Mastery Guide2 gün önce

sovereign AI is the new arms race fr

Gabriel Hernández profil fotoğrafı

Gabriel Hernández2 gün önce

I try to make the notification useful when it arrived. For me, recording why an alert was set at a level helps.

icefrog.◎ profil fotoğrafı

icefrog.◎2 gün önce

good luck auditing weights in a cold war mindset lol

Rosetina Degen 🚢🎒 profil fotoğrafı

Rosetina Degen 🚢🎒2 gün önce

Sovereign AI demands careful evaluation

Sridhar A profil fotoğrafı

Sridhar A2 gün önce

ai needs verification, not just trust. sovereign ai makes that impossible to ignore.

Benzer Videolar

Vals AI CEO Rayan Krishnan on how they score a model's ability to build its own successor: "One benchmark we released recently is our Recursive Self-Improvement Index. It's a topic which a lot of the big labs have been talking about and starting to report on in their model cards." "There isn't a shared language to talk about the RSI potential of models. So we created this as an apples-to-apples way to actually benchmark across the models." "In an ideal world, what you want to do is take a frontier model and have it train the next version of itself and see where the delta comes from. But obviously that's very expensive and slow." "What we're doing is forming a set of proxies for every part of the process it takes to build the next version of the model." "There's work around pre-training, post-training, harness-level engineering, and then seeing in which mechanisms and behaviors the models are able to do very good research work and build something new, and where they're struggling." Rayan Krishnan Jennifer Li

Vals AI CEO Rayan Krishnan on how they score a model's ability to build its own successor: "One benchmark we released recently is our Recursive Self-Improvement Index. It's a topic which a lot of the big labs have been talking about and starting to report on in their model cards." "There isn't a shared language to talk about the RSI potential of models. So we created this as an apples-to-apples way to actually benchmark across the models." "In an ideal world, what you want to do is take a frontier model and have it train the next version of itself and see where the delta comes from. But obviously that's very expensive and slow." "What we're doing is forming a set of proxies for every part of the process it takes to build the next version of the model." "There's work around pre-training, post-training, harness-level engineering, and then seeing in which mechanisms and behaviors the models are able to do very good research work and build something new, and where they're struggling." Rayan Krishnan Jennifer Li

a16z

34,164 görüntüleme • 2 gün önce

🚨 NOW: Vice President Vance SHOOTS DOWN the fake news that Trump will just give away Iran everything without verifying or enforcing anything "We don't trust WORDS, we trust ACTION and we trust conduct — and so we're going to reward conduct and we're not going to reward any words whether they're written on a sheet of paper or not!" "There's a lot of discussion, the MOU, the gentleman's agreements, the final deal. Words don't matter, ladies and gentlemen, we're about verification." "And so what we're going to do is to say, if they do the things that they have promised to do, they have promised not to enrich. They have promised that they would allow inspectors in to destroy that highly enriched stockpile and then of course it's not usable anymore." "They promise a number of things and that's why the deal contemplates a number of benefits if they do those things. But it doesn't do anything if they don't actually meet those promises."

🚨 NOW: Vice President Vance SHOOTS DOWN the fake news that Trump will just give away Iran everything without verifying or enforcing anything "We don't trust WORDS, we trust ACTION and we trust conduct — and so we're going to reward conduct and we're not going to reward any words whether they're written on a sheet of paper or not!" "There's a lot of discussion, the MOU, the gentleman's agreements, the final deal. Words don't matter, ladies and gentlemen, we're about verification." "And so what we're going to do is to say, if they do the things that they have promised to do, they have promised not to enrich. They have promised that they would allow inspectors in to destroy that highly enriched stockpile and then of course it's not usable anymore." "They promise a number of things and that's why the deal contemplates a number of benefits if they do those things. But it doesn't do anything if they don't actually meet those promises."

Eric Daugherty

15,117 görüntüleme • 2 ay önce

DAVID SACKS: “I understand that there is a lot of concern about [AI] job disruption or job loss. That's not what we're seeing in the data so far. What we're actually seeing is job creation … We have a 4% GDP growth number right now, and half of that is attributable to AI.”

DAVID SACKS: “I understand that there is a lot of concern about [AI] job disruption or job loss. That's not what we're seeing in the data so far. What we're actually seeing is job creation … We have a 4% GDP growth number right now, and half of that is attributable to AI.”

Chief Nerd

206,382 görüntüleme • 8 ay önce

Vals AI CEO Rayan Krishnan on why defensible evals are critical for enterprises' survival: "The labs side of it is very clear. If you're raising lots of money, investing heavily in building models, it's essential for you to show why your model is getting better and why the customer should pay a premium for them." "But what I think is still underappreciated is, on the enterprise side, this is turning out to be existential as well." "We're in this world where it is still very unclear what ROI looks like and how to value this intelligence that's being used... Token spend may start to eclipse salary spend." "If this is such a meaningful line item in your costs, you have to justify the ROI much more cleanly... Over time, I think a firm really is just its evals." "The ability for a company to make its evals legible in order to solve this ROI calculus is going to be the reason why that company wins out over its competitors in the long term." Rayan Krishnan Erik Torenberg

Vals AI CEO Rayan Krishnan on why defensible evals are critical for enterprises' survival: "The labs side of it is very clear. If you're raising lots of money, investing heavily in building models, it's essential for you to show why your model is getting better and why the customer should pay a premium for them." "But what I think is still underappreciated is, on the enterprise side, this is turning out to be existential as well." "We're in this world where it is still very unclear what ROI looks like and how to value this intelligence that's being used... Token spend may start to eclipse salary spend." "If this is such a meaningful line item in your costs, you have to justify the ROI much more cleanly... Over time, I think a firm really is just its evals." "The ability for a company to make its evals legible in order to solve this ROI calculus is going to be the reason why that company wins out over its competitors in the long term." Rayan Krishnan Erik Torenberg

a16z

34,017 görüntüleme • 2 gün önce

"My biggest worry is... the perception of technology in America." Ben Horowitz on why optimism about AI is critical: "Over 70% of people in China are optimistic about AI. And less than 30% in America or something like that were optimistic about AI." "We're very focused on the dangers of AI here and not on the positive things." "We're going to... end traffic deaths. We're going to cure cancer. We're going to end poverty as we know it." "Those are pretty good things." "The positive uses of the technology are extremely positive. The negative ones are there, but we can manage them, in the same way we've managed the negative of every technology." "Starting with fire, which has very bad consequences — it can burn down your village... but it also heats your house, and cooks your food, and does some very wonderful things." benahorowitz.eth with David Ulevitch 🇺🇸

"My biggest worry is... the perception of technology in America." Ben Horowitz on why optimism about AI is critical: "Over 70% of people in China are optimistic about AI. And less than 30% in America or something like that were optimistic about AI." "We're very focused on the dangers of AI here and not on the positive things." "We're going to... end traffic deaths. We're going to cure cancer. We're going to end poverty as we know it." "Those are pretty good things." "The positive uses of the technology are extremely positive. The negative ones are there, but we can manage them, in the same way we've managed the negative of every technology." "Starting with fire, which has very bad consequences — it can burn down your village... but it also heats your house, and cooks your food, and does some very wonderful things." benahorowitz.eth with David Ulevitch 🇺🇸

a16z

51,056 görüntüleme • 6 ay önce

"We're about to see the explosion of analog." Gary Vaynerchuk wants to open a restaurant that makes you check your phone in at the door and seats you at communal tables. "Extreme AI is creating extreme analog. I think it's a barbell." "I could not be more interested in physical retail, event-driven businesses, in concerts and venues." "There are a lot of interesting non-digital realities that are coming as a countermove to the insanity of AI advancements." "We're literally within a half decade of not believing a single video that's on the internet. In 5 years, if we're having this interview, most of the audience is trying to figure out if we're real or not." "That is very real, and has substantial counter-opportunities." "Any real entrepreneur, they're not crying about AI killing them. They're curious about how AI at scale is going to create opportunity for them."

"We're about to see the explosion of analog." Gary Vaynerchuk wants to open a restaurant that makes you check your phone in at the door and seats you at communal tables. "Extreme AI is creating extreme analog. I think it's a barbell." "I could not be more interested in physical retail, event-driven businesses, in concerts and venues." "There are a lot of interesting non-digital realities that are coming as a countermove to the insanity of AI advancements." "We're literally within a half decade of not believing a single video that's on the internet. In 5 years, if we're having this interview, most of the audience is trying to figure out if we're real or not." "That is very real, and has substantial counter-opportunities." "Any real entrepreneur, they're not crying about AI killing them. They're curious about how AI at scale is going to create opportunity for them."

TBPN

193,770 görüntüleme • 2 ay önce

"We're about to see the explosion of analog." Gary Vaynerchuk wants to open a restaurant that makes you check your phone in at the door and seats you at communal tables. "Extreme AI is creating extreme analog. I think it's a barbell." "I could not be more interested in physical retail, event-driven businesses, in concerts and venues." "There are a lot of interesting non-digital realities that are coming as a countermove to the insanity of AI advancements." "We're literally within a half decade of not believing a single video that's on the internet. In 5 years, if we're having this interview, most of the audience is trying to figure out if we're real or not." "That is very real, and has substantial counter-opportunities." "Any real entrepreneur, they're not crying about AI killing them. They're curious about how AI at scale is going to create opportunity for them."

"We're about to see the explosion of analog." Gary Vaynerchuk wants to open a restaurant that makes you check your phone in at the door and seats you at communal tables. "Extreme AI is creating extreme analog. I think it's a barbell." "I could not be more interested in physical retail, event-driven businesses, in concerts and venues." "There are a lot of interesting non-digital realities that are coming as a countermove to the insanity of AI advancements." "We're literally within a half decade of not believing a single video that's on the internet. In 5 years, if we're having this interview, most of the audience is trying to figure out if we're real or not." "That is very real, and has substantial counter-opportunities." "Any real entrepreneur, they're not crying about AI killing them. They're curious about how AI at scale is going to create opportunity for them."

TBPN

319,450 görüntüleme • 4 ay önce

Cowboy CEO Baiju Bhatt on why orbital data centers won't just be a SpaceX monopoly. "We're working on it. I think that if there's enough commercial demand for this, and there are enough dollars flowing to this, I think that capitalism is efficient, right?" "I think that as we look at, at least in the US, data centers are not very popular. They're actually quite unpopular, right? "And the intersection between these massive build outs, as we look out to the amount of capacity that people want—we're still kind of the early innings of this stuff." "The frustration from the US population, I see this probably getting worse before it gets better, which means that if AI wants to eat a ton of energy, it's gonna need some other places to get it." "So I think there is a degree of inevitability, not just to data centers in space, but also the other ways of powering AI." "And I think if there's enough commercial activity, you're gonna see a lot of people try to do it. And like I said, we're also doing it." With Mo Islam at the NYSE 🏛

Cowboy CEO Baiju Bhatt on why orbital data centers won't just be a SpaceX monopoly. "We're working on it. I think that if there's enough commercial demand for this, and there are enough dollars flowing to this, I think that capitalism is efficient, right?" "I think that as we look at, at least in the US, data centers are not very popular. They're actually quite unpopular, right? "And the intersection between these massive build outs, as we look out to the amount of capacity that people want—we're still kind of the early innings of this stuff." "The frustration from the US population, I see this probably getting worse before it gets better, which means that if AI wants to eat a ton of energy, it's gonna need some other places to get it." "So I think there is a degree of inevitability, not just to data centers in space, but also the other ways of powering AI." "And I think if there's enough commercial activity, you're gonna see a lot of people try to do it. And like I said, we're also doing it." With Mo Islam at the NYSE 🏛

Jack Kuhr

125,101 görüntüleme • 14 gün önce

Marc Andreessen says AI is hyper democratizing. “Is the future of the world going to be one or a small number of companies, or for that matter, governments or super AIs that kind of own and control everything?” “Or is this a technology that's going to diffuse out and be in everybody's hands?” “This is the time in which AI is actually hyper democratizing. It has spread to more individuals, both in the country and around the world in the shortest period of time of any new technology I think in history.” “The best AIs in the world are in the consumer products.” “This technology is going to be in everybody's hands. And everybody is going to be able to use it to optimize the things that they do, have it be a thought partner, assistant for building companies.” Marc Andreessen 🇺🇸

Marc Andreessen says AI is hyper democratizing. “Is the future of the world going to be one or a small number of companies, or for that matter, governments or super AIs that kind of own and control everything?” “Or is this a technology that's going to diffuse out and be in everybody's hands?” “This is the time in which AI is actually hyper democratizing. It has spread to more individuals, both in the country and around the world in the shortest period of time of any new technology I think in history.” “The best AIs in the world are in the consumer products.” “This technology is going to be in everybody's hands. And everybody is going to be able to use it to optimize the things that they do, have it be a thought partner, assistant for building companies.” Marc Andreessen 🇺🇸

a16z

79,935 görüntüleme • 10 ay önce

From the archive: Brian Armstrong on AI trading bots "We're going to get AI agents that can be your registered investment advisors." "Imagine you are in the Coinbase 🛡️ app and you tell them: I want to put in $10,000, go make me money." "It may make mistakes and there's a lot of details to figure out, but I think having a trading bot do a lot of that for you (is very cool)."

From the archive: Brian Armstrong on AI trading bots "We're going to get AI agents that can be your registered investment advisors." "Imagine you are in the Coinbase 🛡️ app and you tell them: I want to put in $10,000, go make me money." "It may make mistakes and there's a lot of details to figure out, but I think having a trading bot do a lot of that for you (is very cool)."

TBPN

66,252 görüntüleme • 1 yıl önce

President Trump says he and Putin talked about "de-nuclearizing" and that "there's no reason for us to be building brand new nuclear weapons." "We were talking about de-nuking; president Putin and I agreed we are going to do it in a very big way. There's no reason for us to be building brand new nuclear weapons. We already have so many you could destroy the world 50 times over, 100 times over. And here we are building new nuclear weapons and they're building nuclear weapons and China is building nuclear weapons and China is trying to catch up because you know, they're very substantially behind, but ithin five or six years they'll be even. And we're all spending a lot of money that we could be spending on other things that are actually hopefully much more productive. Hopefully there'll never be a time wren with -- when we need those weapons. If there's ever a time when we need nuclear weapon, the kind of weapons that we're building and Russia has and China has to a lesser extent but will have. That's going to be a very sad day. That's going to be probably oblivion."

Ken Klippenstein (NSPM-7 Compliant)

683,841 görüntüleme • 1 yıl önce

David from USD.AI on the institutional shift happening behind CHIP, why every borrower they have is institutional, and why the AI super cycle has had no real on-chain exposure until now: "We have personally beat out other investment banks in running deals that we've signed. They're starting to view us as a threat. One call we had, she ended it early: 'I think you're a competitor, I can't talk anymore.' We're just cheaper financing, even for the bigger deals. People think we only do small deals, but we're starting to get bigger and bigger. As we attract more capital, we have to deploy in bigger deals to keep up. All of our customers are institutional. There's not a single retail borrower. These chips are expensive: The minimum we lend against is H200s, clusters of eight, that's like 300k. We're working with many of the brand names in San Francisco that are actually in the AI sector. People appreciate the positioning of Chip; it's actually engaging with the AI economy. A lot of people claim to be AI. They're in their own version of AI that they think is AI. But then there's the OpenAIs of the world. There's the Cursors of the world. Real customers, with real backing, that have cash flows. That is what people in our space want exposure to. We just haven't had a good way to do it. There's a reason you can get a loan against an airplane or a Honda Civic, but not a GPU. And GPUs are going to be worth a lot more than cars or airplanes in the future." Clip from the Modern Market interview with David | www.usd.ai

The Modern Market Show

10,397 görüntüleme • 4 ay önce

Sam Altman says AI leaders have done a terrible job at communicating to the public: “The, ‘Dear peasants, we will bequeath upon you these gifts of a cure for cancer and material wealth and great entertainment, and you stop complaining, and we'll make all the decisions about the future and just trust us, we'll be benevolent dictators.’ That’s not good. Not good.” “A lot of the people building AI have been saying, ‘There's a 25% chance we're going to destroy the world, and yet we're going to race ahead to do it because otherwise those bad guys will do it first.’” “Or, ‘Man, this thing is going to be really terrible, and 50% of the jobs are going to go away in the next year, and we hope you all are okay, but it seems really scary.’” “We have not, as a field, done a very good job of explaining to people what the benefits are and how the downsides can be mitigated—and we certainly have not done a good job even if people have had answers like saying, ‘There's going to be UBI, or work will be optional.’” “There's been very little discussion from people about how and why it's important that people have more power and personal freedom in the world, not less. And that matters a lot to most people.” “The ability of people to influence their own future and collectively design where society is going to go, and the autonomy that comes with that, is very important. I don't think a lot of people in the AI field… they feel it for themselves, but they don't spend much time thinking about it, reflecting on or acknowledging how important that is to other people.”

Sam Altman says AI leaders have done a terrible job at communicating to the public: “The, ‘Dear peasants, we will bequeath upon you these gifts of a cure for cancer and material wealth and great entertainment, and you stop complaining, and we'll make all the decisions about the future and just trust us, we'll be benevolent dictators.’ That’s not good. Not good.” “A lot of the people building AI have been saying, ‘There's a 25% chance we're going to destroy the world, and yet we're going to race ahead to do it because otherwise those bad guys will do it first.’” “Or, ‘Man, this thing is going to be really terrible, and 50% of the jobs are going to go away in the next year, and we hope you all are okay, but it seems really scary.’” “We have not, as a field, done a very good job of explaining to people what the benefits are and how the downsides can be mitigated—and we certainly have not done a good job even if people have had answers like saying, ‘There's going to be UBI, or work will be optional.’” “There's been very little discussion from people about how and why it's important that people have more power and personal freedom in the world, not less. And that matters a lot to most people.” “The ability of people to influence their own future and collectively design where society is going to go, and the autonomy that comes with that, is very important. I don't think a lot of people in the AI field… they feel it for themselves, but they don't spend much time thinking about it, reflecting on or acknowledging how important that is to other people.”

David Senra

32,300 görüntüleme • 19 gün önce

.Scott Wu to Joe Lonsdale on why Cognition uses AI in job interviews: "One or two years ago, people had this mentality of 'How do we make sure people aren't using AI while we're interviewing them?' And I think that has totally flipped. Honestly, I think that was wrong... If you're asking the question of like, how do we evaluate people on exactly the thing that AI can already do, that seems kind of the wrong. Our interview process has always been... you can use as much AI as you want to use. We're going to give you a few hours. Just build your whole own product surface. A lot of these are projects that, frankly, is like, if you were trying to do this by hand, you would not be able to get done in a few hours. So you kind of have to use AI for this. But in reality, what we actually want to test is, in addition to how familiar you are with these tools, what we actually want to test is, what do you think is the right thing to build? Or how do you make these product decisions? How do you make these trade-offs? How do you decide or collect information about what you should be doing?"

.Scott Wu to Joe Lonsdale on why Cognition uses AI in job interviews: "One or two years ago, people had this mentality of 'How do we make sure people aren't using AI while we're interviewing them?' And I think that has totally flipped. Honestly, I think that was wrong... If you're asking the question of like, how do we evaluate people on exactly the thing that AI can already do, that seems kind of the wrong. Our interview process has always been... you can use as much AI as you want to use. We're going to give you a few hours. Just build your whole own product surface. A lot of these are projects that, frankly, is like, if you were trying to do this by hand, you would not be able to get done in a few hours. So you kind of have to use AI for this. But in reality, what we actually want to test is, in addition to how familiar you are with these tools, what we actually want to test is, what do you think is the right thing to build? Or how do you make these product decisions? How do you make these trade-offs? How do you decide or collect information about what you should be doing?"

American Optimist

13,182 görüntüleme • 5 ay önce

"Enterprise makes a lot of sense until everyone does enterprise." Brian Chesky says we're going to see a massive revolution from enterprise to consumer AI in the next two years: "I saw a tweet recently that like 60 new neolabs are being formed. I've met a number of these researchers. A lot of them are interested in doing the same things." "Everyone's saying that consumer's hard. But you know what's also hard? Competing with 10 other companies in enterprise doing your idea too. That's also hard." "We're about to enter this new era where suddenly going beyond text is possible. I think you're going to even be able to see real-time AI-generated interfaces in the next year or two." "My prediction is, this has been the era of enterprise. Even [with] ChatGPT's breakout success, I would predict that most of their revenue is going to be going to Codex. And I think in the next two years, you're going to see a massive revolution in consumer." "We're going to try to do our part. I'm not sure Airbnb will be the company to blaze the trail. But I think someone has to. In another era, it would have been Apple. The Steve Jobs-era Apple would have done that."

"Enterprise makes a lot of sense until everyone does enterprise." Brian Chesky says we're going to see a massive revolution from enterprise to consumer AI in the next two years: "I saw a tweet recently that like 60 new neolabs are being formed. I've met a number of these researchers. A lot of them are interested in doing the same things." "Everyone's saying that consumer's hard. But you know what's also hard? Competing with 10 other companies in enterprise doing your idea too. That's also hard." "We're about to enter this new era where suddenly going beyond text is possible. I think you're going to even be able to see real-time AI-generated interfaces in the next year or two." "My prediction is, this has been the era of enterprise. Even [with] ChatGPT's breakout success, I would predict that most of their revenue is going to be going to Codex. And I think in the next two years, you're going to see a massive revolution in consumer." "We're going to try to do our part. I'm not sure Airbnb will be the company to blaze the trail. But I think someone has to. In another era, it would have been Apple. The Steve Jobs-era Apple would have done that."

TBPN

24,882 görüntüleme • 4 ay önce

Demis on why world models are his longest standing passion and explains benefits vs. language models: ▫️ “I think language models are able to understand a lot about the world. More than we expected because language is actually probably richer than we thought. But there's still a lot about the spatial dynamics of the world, spatial awareness and the physical context we're in — and how that works mechanically — that is hard to describe in words and isn't generally described in corpuses of words. Alot of this is allied to learning from experience. There's a lot of things which you can't really describe something. You have to just experience it. Maybe the senses and so on are very hard to put into words. Whether that's motor angles and smell and these kinds of senses, it's very difficult to describe that in any kind of language.”▫️

Demis on why world models are his longest standing passion and explains benefits vs. language models: ▫️ “I think language models are able to understand a lot about the world. More than we expected because language is actually probably richer than we thought. But there's still a lot about the spatial dynamics of the world, spatial awareness and the physical context we're in — and how that works mechanically — that is hard to describe in words and isn't generally described in corpuses of words. Alot of this is allied to learning from experience. There's a lot of things which you can't really describe something. You have to just experience it. Maybe the senses and so on are very hard to put into words. Whether that's motor angles and smell and these kinds of senses, it's very difficult to describe that in any kind of language.”▫️

Bearly AI

78,545 görüntüleme • 3 ay önce

We asked Isaiah Taylor - making nuclear reactors about the nuclear announcements today. "There hasn't been a Ford or Tesla of nuclear energy yet. "Valor Atomics intends to be that company." "If we're going to win on AI, manufacturing, or supply chain, we're going to do it with a ton of cheap energy."

We asked Isaiah Taylor - making nuclear reactors about the nuclear announcements today. "There hasn't been a Ford or Tesla of nuclear energy yet. "Valor Atomics intends to be that company." "If we're going to win on AI, manufacturing, or supply chain, we're going to do it with a ton of cheap energy."

TBPN

23,539 görüntüleme • 1 yıl önce

Demis Hassabis says world models are his longest standing passion and explains benefits vs. language models: ▫️ “I think language models are able to understand a lot about the world. More than we expected because language is actually probably richer than we thought. But there's still a lot about the spatial dynamics of the world, spatial awareness and the physical context we're in — and how that works mechanically — that is hard to describe in words and isn't generally described in corpuses of words. Alot of this is allied to learning from experience. There's a lot of things which you can't really describe something. You have to just experience it. Maybe the senses and so on are very hard to put into words. Whether that's motor angles and smell and these kinds of senses, it's very difficult to describe that in any kind of language.”▫️ This is what Demis and Google Deepmind is trying to solve with Genie. He also says that the video models (Veo) will play a part in training the world models and this is all key for AI robotics.

Demis Hassabis says world models are his longest standing passion and explains benefits vs. language models: ▫️ “I think language models are able to understand a lot about the world. More than we expected because language is actually probably richer than we thought. But there's still a lot about the spatial dynamics of the world, spatial awareness and the physical context we're in — and how that works mechanically — that is hard to describe in words and isn't generally described in corpuses of words. Alot of this is allied to learning from experience. There's a lot of things which you can't really describe something. You have to just experience it. Maybe the senses and so on are very hard to put into words. Whether that's motor angles and smell and these kinds of senses, it's very difficult to describe that in any kind of language.”▫️ This is what Demis and Google Deepmind is trying to solve with Genie. He also says that the video models (Veo) will play a part in training the world models and this is all key for AI robotics.

Bearly AI

123,703 görüntüleme • 8 ay önce

Demis Hassabis on why world models are his longest standing passion and the benefits vs. language models: ▫️ “I think language models are able to understand a lot about the world. More than we expected because language is actually probably richer than we thought. But there's still a lot about the spatial dynamics of the world, spatial awareness and the physical context we're in — and how that works mechanically — that is hard to describe in words and isn't generally described in corpuses of words. Alot of this is allied to learning from experience. There's a lot of things which you can't really describe something. You have to just experience it. Maybe the senses and so on are very hard to put into words. Whether that's motor angles and smell and these kinds of senses, it's very difficult to describe that in any kind of language.”▫️ *** From the Google DeepMind podcast in December 2025.

Demis Hassabis on why world models are his longest standing passion and the benefits vs. language models: ▫️ “I think language models are able to understand a lot about the world. More than we expected because language is actually probably richer than we thought. But there's still a lot about the spatial dynamics of the world, spatial awareness and the physical context we're in — and how that works mechanically — that is hard to describe in words and isn't generally described in corpuses of words. Alot of this is allied to learning from experience. There's a lot of things which you can't really describe something. You have to just experience it. Maybe the senses and so on are very hard to put into words. Whether that's motor angles and smell and these kinds of senses, it's very difficult to describe that in any kind of language.”▫️ *** From the Google DeepMind podcast in December 2025.

Bearly AI

65,630 görüntüleme • 1 ay önce