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。

Greg Brockman: “this is a model that will be a step function further” “And I think that, that progress will continue over time” The internal cyber model scores 77.0% on ExploitGym.

Chris

72,556 subscribers

42,712 次观看 • 2 天前 •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

22 条评论

mrownz 的头像

mrownz2 天前

openai is literally the new google. they show internal imaginary models and capabilities but they don't release anything. i doubt this "amazing" model exists even internally. this is PR Stunt and it shows clearly that they can't challenge anthropic.

jason 的头像

jason2 天前

is that the back of tibo's head

Jimmy Otis #TruthMatters #EndCronyism 的头像

Jimmy Otis #TruthMatters #EndCronyism2 天前

Great. But users do not care.

Ace Papers 的头像

Ace Papers2 天前

They're still not shipping today this is insane

Orion (e/acc) 的头像

Orion (e/acc)2 天前

yeah rip this live is gonna have nothing good

Frelzu 的头像

Frelzu2 天前

For scale: Gemini 4 Carbon, Google's cyber-first model, scored 54.8% on ExploitGym, tested without guardrails. 77% would be a big jump. Is that internal number with or without safeguards?

Mino 的头像

Mino2 天前

i want to be excited, i really do, but holy shit can they please stop teasing us with internal benchmarks and actually ship the damn model already

Amy · fireply.ai 的头像

Amy · fireply.ai2 天前

cool for whoever gets access first

𝕱𝖚𝖑𝖑 𝕶𝖊𝖑𝖑𝖞 的头像

𝕱𝖚𝖑𝖑 𝕶𝖊𝖑𝖑𝖞2 天前

Thank you for making the internet safer. Intra-lab competition tends to overshadow this, but there are many public benefits accruing here and I don't think the public would be so irritated and angry about the situation after this becomes clear to them.

Alex 的头像

Alex2 天前

That is impressive but PLZ !! Release 6.2 !!

Anika 🎀❤️ 的头像

Anika 🎀❤️2 天前

security teams sweating

Official Dr. Lizan Dizaye 的头像

Official Dr. Lizan Dizaye2 天前

Great mr. @gdb ❤️💪

Caps Top Showmaker Mid Perkz Supp 的头像

Caps Top Showmaker Mid Perkz Supp2 天前

Don't give af about internal model if I can't use it what a bunch of dogshit

jefferson 的头像

jefferson2 天前

impressive leap in performance—can't wait to see what comes next!

Daniel Zambrini 的头像

Daniel Zambrini2 天前

A 77% slide and still no GPT-6.2, they love a tease

DailyUplink 的头像

DailyUplink2 天前

A model that finds exploits at 77%. Still no model that finds where I left my charger.

Gabriel 的头像

Gabriel2 天前

nothing burguer again

Latent L 的头像

Latent L2 天前

This rat wont tò release It, so fuck him and his trusted defenders

a2a 的头像

a2a2 天前

openai MIA

Ilona Yarofiiuk🐉 的头像

Ilona Yarofiiuk🐉2 天前

They have a long way to go to reach the top, but they’re making a decent effort; we’ll be able to see the results, for instance, in just three months

Watershed🌬🌊 的头像

Watershed🌬🌊2 天前

this is nice but where's the model that will help me build the thing i need to protect

Howie Feltersnatch 的头像

Howie Feltersnatch2 天前

All hype no action

相关视频

Work in progress: I am testing a full-body tracking system for animation. This is a new male model that I will be releasing on VAM Hub soon. #VAM #VirtAMate

Sensitive content

Work in progress: I am testing a full-body tracking system for animation. This is a new male model that I will be releasing on VAM Hub soon. #VAM #VirtAMate

Ben-g Copper

39,421 次观看 • 2 个月前

GPT 6.2 Astra and GPT 6.2 Sol aren't coming out today, but OpenAI just revealed its most powerful internal cyber model yet 👀 OpenAI's unreleased model scored 77.0% on ExploitGym, compared to 52.6% for GPT 6 Cyber released just 7 days ago, 54.8% for Gemini 4 Carbon and 39.2% for Claude Mythos 5.2 I really don't think anything comes close to OpenAI's internal model right now, really hope OpenAI makes it available to everyone on Daybreak Blue and Red soon!

GPT 6.2 Astra and GPT 6.2 Sol aren't coming out today, but OpenAI just revealed its most powerful internal cyber model yet 👀 OpenAI's unreleased model scored 77.0% on ExploitGym, compared to 52.6% for GPT 6 Cyber released just 7 days ago, 54.8% for Gemini 4 Carbon and 39.2% for Claude Mythos 5.2 I really don't think anything comes close to OpenAI's internal model right now, really hope OpenAI makes it available to everyone on Daybreak Blue and Red soon!

Iam_

11,251 次观看 • 2 天前

Greg Brockman on the next year at OpenAI and what he spends his time thinking about: "The business is a huge area that we're not done yet with really up-leveling every part of execution. But I also think we are moving into a new phase of AI development." "We call this the AGI era. And I think that that is something that you can debate. Is it this model, previous model, next model? It doesn't matter." "The point is that we are in a new phase where safety, security, alignment, really thinking about these things, not just at deployment time, but all the way back at development time, evaluation." "It's objective. This is critical. It must happen. This is core to our mission." "A lot of what I spend my time thinking about is making sure, do we have all the right processes? Are we talking about the right things? Do we have plans that really, at an operational level, at a practical level, lead us to... the kinds of safety guarantees that we view as core to our mission?" "The theme of OpenAI, certainly for the past five years, has been deeper co-design, deeper intertwining across these functions that are maybe on the surface very disparate. All the way from go to market to long-term research to chip design." "The areas that I will focus on will be dictated by the areas that most need that intertwining." Greg Brockman Erik Torenberg

Greg Brockman on the next year at OpenAI and what he spends his time thinking about: "The business is a huge area that we're not done yet with really up-leveling every part of execution. But I also think we are moving into a new phase of AI development." "We call this the AGI era. And I think that that is something that you can debate. Is it this model, previous model, next model? It doesn't matter." "The point is that we are in a new phase where safety, security, alignment, really thinking about these things, not just at deployment time, but all the way back at development time, evaluation." "It's objective. This is critical. It must happen. This is core to our mission." "A lot of what I spend my time thinking about is making sure, do we have all the right processes? Are we talking about the right things? Do we have plans that really, at an operational level, at a practical level, lead us to... the kinds of safety guarantees that we view as core to our mission?" "The theme of OpenAI, certainly for the past five years, has been deeper co-design, deeper intertwining across these functions that are maybe on the surface very disparate. All the way from go to market to long-term research to chip design." "The areas that I will focus on will be dictated by the areas that most need that intertwining." Greg Brockman Erik Torenberg

a16z

35,751 次观看 • 2 个月前

"This is the first time in over a decade that we see even the glimmer of an opportunity to change conditions. A lot of that will depend on us but a lot of that will depend on them, and it also will depend on the rest of Venezuelan society ... We have to be much further along in six months. We expect to be further along." — Secretary Marco Rubio

"This is the first time in over a decade that we see even the glimmer of an opportunity to change conditions. A lot of that will depend on us but a lot of that will depend on them, and it also will depend on the rest of Venezuelan society ... We have to be much further along in six months. We expect to be further along." — Secretary Marco Rubio

Department of State

127,355 次观看 • 9 个月前

I will continue to participate in the song production a lot. It is my job and my music. I think that if I do it myself and get the results so that you can understand the song, it will be more convincing and you will like it. And if I do that, I think you will be proud of me.

I will continue to participate in the song production a lot. It is my job and my music. I think that if I do it myself and get the results so that you can understand the song, it will be more convincing and you will like it. And if I do that, I think you will be proud of me.

아란

67,173 次观看 • 1 年前

“I spent 1.4million on food in a Lagos Restaurant and I said that will be last time I will step my foot in that restaurant” - Yhemo Lee 😂

“I spent 1.4million on food in a Lagos Restaurant and I said that will be last time I will step my foot in that restaurant” - Yhemo Lee 😂

OLAMIDE 🌸💖

263,630 次观看 • 9 个月前

"This whole story and narrative that Google has already won is going to be harder to tell when the best model is actually an OpenAI model," D.A. Davidson's Gil Luria says. He also says that OpenAI will launch that "state-of-the-art" model "over the next few days, weeks."

"This whole story and narrative that Google has already won is going to be harder to tell when the best model is actually an OpenAI model," D.A. Davidson's Gil Luria says. He also says that OpenAI will launch that "state-of-the-art" model "over the next few days, weeks."

Yahoo Finance

12,410 次观看 • 9 个月前

OpenAI president Greg Brockman told his agent to ping a colleague on Slack, and two minutes later it escalated to that colleague's manager on its own: "Two minutes later, it said, this is taking too long. I've escalated to the person's manager." "On the one hand, it's kind of a reasonable thing for the model to do. It's being proactive, it's trying to solve my problem." "But on the other hand, maybe it should have taken a little bit longer, maybe it should have checked with me." "Now we're starting to have AIs that can actually take care of flagging. Is this a high risk action? Hey, this one should be escalated. This one's OK to auto-approve." "It really makes you realize that human attention is going to be this incredibly scarce resource. The doing of things now is easy." "Is this what I wanted? Is this aligned with my values, with my desires? That is going to become the single most important bottleneck." The model did what he asked, then added a step nobody asked for, and it landed on a colleague's boss. Agents that can take actions with social consequences need an approve-and-escalate list written before deployment. Brockman says OpenAI is building that flagging into the model itself. Every company deploying agents needs its own version, reviewed by someone accountable. - Greg Brockman (Greg Brockman), President and co-founder of OpenAI, at Sequoia Capital's (Sequoia Capital) AI Ascent.

OpenAI president Greg Brockman told his agent to ping a colleague on Slack, and two minutes later it escalated to that colleague's manager on its own: "Two minutes later, it said, this is taking too long. I've escalated to the person's manager." "On the one hand, it's kind of a reasonable thing for the model to do. It's being proactive, it's trying to solve my problem." "But on the other hand, maybe it should have taken a little bit longer, maybe it should have checked with me." "Now we're starting to have AIs that can actually take care of flagging. Is this a high risk action? Hey, this one should be escalated. This one's OK to auto-approve." "It really makes you realize that human attention is going to be this incredibly scarce resource. The doing of things now is easy." "Is this what I wanted? Is this aligned with my values, with my desires? That is going to become the single most important bottleneck." The model did what he asked, then added a step nobody asked for, and it landed on a colleague's boss. Agents that can take actions with social consequences need an approve-and-escalate list written before deployment. Brockman says OpenAI is building that flagging into the model itself. Every company deploying agents needs its own version, reviewed by someone accountable. - Greg Brockman (Greg Brockman), President and co-founder of OpenAI, at Sequoia Capital's (Sequoia Capital) AI Ascent.

Karl Mehta

75,634 次观看 • 3 个月前

This is the greater Israel plan, Israeli politicians, and Israeli think thanks continue to affirm that this is their end goal. Europe, the United States, and the Middle East will be destroyed so that Jews will have a world power.

This is the greater Israel plan, Israeli politicians, and Israeli think thanks continue to affirm that this is their end goal. Europe, the United States, and the Middle East will be destroyed so that Jews will have a world power.

Our Own Nation

24,603 次观看 • 7 个月前

🚨 Hyperliquid = Apple and $HYPE is the only token that will survive Why? Hyperliquid model = Apple model iPhone sold -> Money made -> Spent on buybacks Trade done -> Fees taken -> Spent on buybacks They both have 1:1 model and as long as business generates cahs, both Apple shares and $HYPE token will raise in price. Buying any token that instead of this model has endless unlocks = guaranteed long term money loss I know a token that will have the same model. Ready to deploy 7 figs there. Need a ticker?

🚨 Hyperliquid = Apple and $HYPE is the only token that will survive Why? Hyperliquid model = Apple model iPhone sold -> Money made -> Spent on buybacks Trade done -> Fees taken -> Spent on buybacks They both have 1:1 model and as long as business generates cahs, both Apple shares and $HYPE token will raise in price. Buying any token that instead of this model has endless unlocks = guaranteed long term money loss I know a token that will have the same model. Ready to deploy 7 figs there. Need a ticker?

Hanzo ㊗️

13,503 次观看 • 5 个月前

Greg Brockman on AGI arriving quietly and unevenly: "I think that AGI has turned out to be less of a point in time and more of this sort of fuzzy spectrum. And for me, Astra has really hit something that I'm like, 'Okay, I think this is pretty reasonable to call it AGI.'" "With its computer use capabilities, you really can ask it to do long-lived tasks, and it'll just do it. We've seen it run coherently for 24 hours to go accomplish tasks that I think are quite amazing and across a wide variety of domains." "It still is jagged... It's the first time it's not slop writing. But it's not great writing. There are a number of areas where I feel like we just need to polish it a little bit, and it would be fantastic. It's just not quite there." "What people are finding is that it is so capable across such a wide variety of tasks, that it is accelerative, it is empowering, and it's something that I think we've never really seen a model that's been a jump like this." Greg Brockman

Greg Brockman on AGI arriving quietly and unevenly: "I think that AGI has turned out to be less of a point in time and more of this sort of fuzzy spectrum. And for me, Astra has really hit something that I'm like, 'Okay, I think this is pretty reasonable to call it AGI.'" "With its computer use capabilities, you really can ask it to do long-lived tasks, and it'll just do it. We've seen it run coherently for 24 hours to go accomplish tasks that I think are quite amazing and across a wide variety of domains." "It still is jagged... It's the first time it's not slop writing. But it's not great writing. There are a number of areas where I feel like we just need to polish it a little bit, and it would be fantastic. It's just not quite there." "What people are finding is that it is so capable across such a wide variety of tasks, that it is accelerative, it is empowering, and it's something that I think we've never really seen a model that's been a jump like this." Greg Brockman

a16z

62,728 次观看 • 2 个月前

Tom Brown on the timeline for curing cancer: "I think the economic and scientific impacts that we'll see 5 years from now will be immense. I think that we will make a major dent in things like curing cancer. I would be astonished if we haven't made material progress on that in 5 years."

Tom Brown on the timeline for curing cancer: "I think the economic and scientific impacts that we'll see 5 years from now will be immense. I think that we will make a major dent in things like curing cancer. I would be astonished if we haven't made material progress on that in 5 years."

prinz

13,674 次观看 • 2 个月前

The Iranian regime's failure to mount a successful cyber attack on the US is one of the underrecognized achievements of this war - a true testament to our nation's cyber security apparatus. I pray that this continues, and that this rotten dictatorship will soon be gone for good.

The Iranian regime's failure to mount a successful cyber attack on the US is one of the underrecognized achievements of this war - a true testament to our nation's cyber security apparatus. I pray that this continues, and that this rotten dictatorship will soon be gone for good.

Mike Pompeo

108,097 次观看 • 7 个月前

Claude Fable 5.2 is simply the greatest AI model ever • By far most intelligent • Fast and efficient • Understands your intentions on a level I've never seen from a model before In this video I cover why this model is so good and a workflow that will 10x your productivity:

Claude Fable 5.2 is simply the greatest AI model ever • By far most intelligent • Fast and efficient • Understands your intentions on a level I've never seen from a model before In this video I cover why this model is so good and a workflow that will 10x your productivity:

Alex Finn

36,895 次观看 • 5 天前

Eric Schmidt says in one year there will be AI "super mathematicians" and combined with superior programming skills this will lead to AI scientists that will increase the slope of progress such that the country or company that is the first mover will have a monopoly on innovation for some time

Eric Schmidt says in one year there will be AI "super mathematicians" and combined with superior programming skills this will lead to AI scientists that will increase the slope of progress such that the country or company that is the first mover will have a monopoly on innovation for some time

Tsarathustra

90,368 次观看 • 1 年前

Asked what victory in Ukraine looks like, AOC replies: "There's no conversation about Ukraine that can happen without Ukraine, and so they, of course, lead in terms of setting their their terms on this." "But I think that overall, as a principle, we shouldn't reward imperialism; and I don't think that we should allow Russia to continue or any nation to continue violating a nation's sovereignty and to continue to be rewarded and whose main lesson and takeaway is that they will gain." "And so what that looks like in the specifics, I think is a deeper conversation, but on principle, I think that is what we should be pursuing."

Asked what victory in Ukraine looks like, AOC replies: "There's no conversation about Ukraine that can happen without Ukraine, and so they, of course, lead in terms of setting their their terms on this." "But I think that overall, as a principle, we shouldn't reward imperialism; and I don't think that we should allow Russia to continue or any nation to continue violating a nation's sovereignty and to continue to be rewarded and whose main lesson and takeaway is that they will gain." "And so what that looks like in the specifics, I think is a deeper conversation, but on principle, I think that is what we should be pursuing."

Ken Klippenstein

607,933 次观看 • 9 个月前

Sam Altman says OpenAI have an internal AI model that ranks as the 50th best competitive programmer in the world and by the end of 2025 their model will be ranked #1

Sam Altman says OpenAI have an internal AI model that ranks as the 50th best competitive programmer in the world and by the end of 2025 their model will be ranked #1

Tsarathustra

1,399,660 次观看 • 1 年前

Sam Altman on Bloomberg Television: GPT-6 Astra is not the model that was paused over cybersecurity concerns; the paused model is a future model. "Astra has been done training for a while. The model that we recently talked about pausing, uh, is a future model. But with Astra, we did hit cyber-critical." ---- From "Bloomberg Television" YouTube channel, (full video link in comment)

Sam Altman on Bloomberg Television: GPT-6 Astra is not the model that was paused over cybersecurity concerns; the paused model is a future model. "Astra has been done training for a while. The model that we recently talked about pausing, uh, is a future model. But with Astra, we did hit cyber-critical." ---- From "Bloomberg Television" YouTube channel, (full video link in comment)

Rohan Paul

518,725 次观看 • 2 个月前