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。

So much discussion about whether we should be using Chinese open-source models in the US. But how could they actually hurt us? Matan Grimberg Matan Grinberg breaks down the biggest risks- 1. Intentional vulnerabilities. A model could contain backdoors, sleeper behavior, or other weaknesses deliberately introduced into the weights,... potentially sitting dormant until the right conditions trigger them. 2. Unintentional vulnerabilities. A model built in China will naturally be trained and optimized around Chinese users, codebases, languages, and data. It could perform incredibly well overall while having subtle blind spots or security weaknesses in Western systems that nobody intentionally put there. The third we may not know for years. The most concerning failures might not show up in today’s benchmarks or security evaluations. If models become deeply embedded across companies and infrastructure, vulnerabilities could emerge only after we’ve become dependent on them.show more

Sabrina Halper

15,865 subscribers

17,244 次观看 • 23 天前 •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

9 条评论

Sabrina Halper 的头像

Sabrina Halper23 天前

Full episode on YouTube:

Droid 的头像

Droid23 天前

@matanSF That's my CEO!

Ashley Jepson 的头像

Ashley Jepson23 天前

@matanSF the ceo aura is palpable

Rick A.F. 的头像

Rick A.F.22 天前

I don't know if this true at all tbh. I use deepseek, I have Codex/OpenAI check it with SOL. I mix up models all the time, they collaborate, they change, they code. I don't think there is any hidden stuff like iin this video. Even sleeper behavior does not make sense to me, these models can be hosted locally and can not act alone?

Vikram Angrish 的头像

Vikram Angrish22 天前

@matanSF Compared to what? Closed source? Total Blackbox?

Sebastian Escarrer 的头像

Sebastian Escarrer22 天前

@matanSF the debate is stuck on whether to use them. companies already are. the real question is whether anyone deploying them can say what's inside

号角的头像

号角22 天前

@matanSF 这是个傻逼吗？开源模型的权重全部都是可以审核的，它的架构推理你也可以部署在本地。比模型权重没有开放，系统提示词不明确的，推理过程也是黑盒的，要好很多吧？

Richa Sharma 的头像

Richa Sharma23 天前

@matanSF The concept of backdoors still fascinates me actually the first time I heard it (was at a party) from @hendrycks . He could be another great podcast guest Sabrina!

sophs b 的头像

sophs b22 天前

@matanSF Intentional backdoors are a plausible risk, but the bigger threat may be subtle, unintentional biases baked into training data that can't be easily audited. Where's the line between acceptable and unacceptable foreign influence?

相关视频

“We have a crisis of open source models in the Western world. Outside of China, there are no good open source models. We don't even have any in the US now. The talent, the capital and the focus to be best-in-class at pre-training, mid-training, post-training is an extremely scarce skillset. The answer for a lot of countries may just be to take an open weights model from China, post-train or fine-tune their own version, and have that be what they start from." Everett Randle Why does the West have such poor open source models and is it a national security threat to rely on open-source Chinese models Demis Hassabis Yann LeCun Aidan Gomez Lilian Weng Jan Leike

“We have a crisis of open source models in the Western world. Outside of China, there are no good open source models. We don't even have any in the US now. The talent, the capital and the focus to be best-in-class at pre-training, mid-training, post-training is an extremely scarce skillset. The answer for a lot of countries may just be to take an open weights model from China, post-train or fine-tune their own version, and have that be what they start from." Everett Randle Why does the West have such poor open source models and is it a national security threat to rely on open-source Chinese models Demis Hassabis Yann LeCun Aidan Gomez Lilian Weng Jan Leike

Harry Stebbings

63,738 次观看 • 3 个月前

The biggest AI companies may be using safety to lock everyone else out of the future (Save this). David Sacks believes that warnings about catastrophic AI risks will build public support for a powerful federal regulator. That regulator may never explicitly ban open source AI. Instead, it could require every advanced model to remain continuously monitored, centrally controlled and capable of being withdrawn. Closed models such as Claude could satisfy those requirements because they operate on company controlled servers. Open weight models would struggle to comply because anyone can download, copy and modify their underlying weights. Once those weights are publicly released, the developer cannot recall every copy, monitor every user or guarantee that its original safeguards remain intact. Anthropic identifies this irreversibility as a legitimate security concern. Applying identical rules to open and closed models could therefore produce very unequal consequences. Anthropic and OpenAI could afford expensive testing, licensing and monitoring requirements, while startups and independent developers might be unable to comply. The eventual result could be a government protected oligopoly dominated by a few closed model companies. There is evidence supporting part of Sacks’ concern. Anthropic advocates mandatory pre-release testing for every sufficiently powerful model, whether it is open or closed, with evaluations focused on cyber, biological and alignment risks. However, Anthropic explicitly denies supporting a blanket ban on open weight models. The company describes open models without dangerous capabilities as a public good and argues that regulation should depend on demonstrated capabilities rather than whether a model is open or closed. Sacks’ strongest argument is therefore about regulatory consequences because safety organizations could receive stronger protections, politicians could acquire greater authority and dominant AI companies could gain an expensive compliance moat. And open source competitors could gradually be eliminated without the government ever formally announcing a ban.

The biggest AI companies may be using safety to lock everyone else out of the future (Save this). David Sacks believes that warnings about catastrophic AI risks will build public support for a powerful federal regulator. That regulator may never explicitly ban open source AI. Instead, it could require every advanced model to remain continuously monitored, centrally controlled and capable of being withdrawn. Closed models such as Claude could satisfy those requirements because they operate on company controlled servers. Open weight models would struggle to comply because anyone can download, copy and modify their underlying weights. Once those weights are publicly released, the developer cannot recall every copy, monitor every user or guarantee that its original safeguards remain intact. Anthropic identifies this irreversibility as a legitimate security concern. Applying identical rules to open and closed models could therefore produce very unequal consequences. Anthropic and OpenAI could afford expensive testing, licensing and monitoring requirements, while startups and independent developers might be unable to comply. The eventual result could be a government protected oligopoly dominated by a few closed model companies. There is evidence supporting part of Sacks’ concern. Anthropic advocates mandatory pre-release testing for every sufficiently powerful model, whether it is open or closed, with evaluations focused on cyber, biological and alignment risks. However, Anthropic explicitly denies supporting a blanket ban on open weight models. The company describes open models without dangerous capabilities as a public good and argues that regulation should depend on demonstrated capabilities rather than whether a model is open or closed. Sacks’ strongest argument is therefore about regulatory consequences because safety organizations could receive stronger protections, politicians could acquire greater authority and dominant AI companies could gain an expensive compliance moat. And open source competitors could gradually be eliminated without the government ever formally announcing a ban.

Milk Road AI

56,257 次观看 • 13 天前

SITUATION EXPLAINED: What happens to open source when the US restricts access to frontier AI models? We asked Adrian Dittmann: "China is releasing these models [open source] to gain international dominance, people will work on these things and say, 'Hey, maybe I can contribute to this if it's open source.' This is the thing that actually gets them talent." "If you're not allowed into the next GPT release or you're not allowed to use Fable, then what are you gonna do? You're stuck with the existing stuff, or you will have to use an open source model that might have subtly better capabilities." "The United States will catch up to open source in some form as well. They seem to just be maximizing product creation." "The problem with open source models is you have to be kind of a nerd in order to use them properly. Consumer facing, the effects will be minimal because the average person doesn't care about open source models. They only care whether or not they can do a quick search with Gemini or ChatGPT."

SITUATION EXPLAINED: What happens to open source when the US restricts access to frontier AI models? We asked Adrian Dittmann: "China is releasing these models [open source] to gain international dominance, people will work on these things and say, 'Hey, maybe I can contribute to this if it's open source.' This is the thing that actually gets them talent." "If you're not allowed into the next GPT release or you're not allowed to use Fable, then what are you gonna do? You're stuck with the existing stuff, or you will have to use an open source model that might have subtly better capabilities." "The United States will catch up to open source in some form as well. They seem to just be maximizing product creation." "The problem with open source models is you have to be kind of a nerd in order to use them properly. Consumer facing, the effects will be minimal because the average person doesn't care about open source models. They only care whether or not they can do a quick search with Gemini or ChatGPT."

MTS

14,406 次观看 • 3 个月前

NVIDIA CEO, Jensen Huang, warns that America's lead in AI is far from secure. He breaks down the US-China AI competition into what he calls a "five-layer cake." And while the US dominates some layers, Jensen sees critical vulnerabilities in others... 1) Energy: China has twice as much as the US despite a smaller economy. Which "makes no sense" to Jensen. 2) Chips: The US is "generations ahead," but Jensen warns against complacency. "Anybody who thinks China can't manufacture is missing the big idea." 3) Infrastructure: Standing up a data center in the US takes about three years. In China? "They can build a hospital in a weekend." 4) Models: US frontier models are "unquestionably world class," but "China is well ahead, way ahead on open source." 5) Applications: Public sentiment differs sharply. Ask both populations whether AI will do more good than harm, and "in their case 80% would say AI will do more good than harm. In our case, it'd be the other way around." Jensen's warning is clear. Leading in chips and frontier models isn't enough when you're behind on energy, infrastructure speed, open source, and public trust. Winning the AI race requires strength across the entire stack, and right now, the US has work to do.

NVIDIA CEO, Jensen Huang, warns that America's lead in AI is far from secure. He breaks down the US-China AI competition into what he calls a "five-layer cake." And while the US dominates some layers, Jensen sees critical vulnerabilities in others... 1) Energy: China has twice as much as the US despite a smaller economy. Which "makes no sense" to Jensen. 2) Chips: The US is "generations ahead," but Jensen warns against complacency. "Anybody who thinks China can't manufacture is missing the big idea." 3) Infrastructure: Standing up a data center in the US takes about three years. In China? "They can build a hospital in a weekend." 4) Models: US frontier models are "unquestionably world class," but "China is well ahead, way ahead on open source." 5) Applications: Public sentiment differs sharply. Ask both populations whether AI will do more good than harm, and "in their case 80% would say AI will do more good than harm. In our case, it'd be the other way around." Jensen's warning is clear. Leading in chips and frontier models isn't enough when you're behind on energy, infrastructure speed, open source, and public trust. Winning the AI race requires strength across the entire stack, and right now, the US has work to do.

Big Brain AI

59,540 次观看 • 9 个月前

Amp CEO, Quinn Slack on why he thinks Kimi K3 could be a Sputnik moment for America: "With GLM 5.2 we had a mini Sputnik moment. When the Soviet Union launched a satellite into space before the US and we realized we have to catch up." "Kimi K3 I need to go and see, but that could be an even bigger Sputnik moment that wakes us up and will probably create a lot of good effects in terms of catalyzing the understanding of AI as a strategic advantage for the US rather than as something people should be against." "But I think it could also scare a lot of people, and I can understand why they'd be scared." "The worst possible outcome could be that open models are banned or are banned from certain industries or companies in the US. There's a lot of people in lobbying organizations in Washington DC, and a lot of people in various parts of the government that want to do that." "They want to say, we won't let these models trained in China be used in the US. There's a lot of people trying to do that now, and that would make us all worse off. I'm pretty nervous about that." Quinn Slack Amp

Amp CEO, Quinn Slack on why he thinks Kimi K3 could be a Sputnik moment for America: "With GLM 5.2 we had a mini Sputnik moment. When the Soviet Union launched a satellite into space before the US and we realized we have to catch up." "Kimi K3 I need to go and see, but that could be an even bigger Sputnik moment that wakes us up and will probably create a lot of good effects in terms of catalyzing the understanding of AI as a strategic advantage for the US rather than as something people should be against." "But I think it could also scare a lot of people, and I can understand why they'd be scared." "The worst possible outcome could be that open models are banned or are banned from certain industries or companies in the US. There's a lot of people in lobbying organizations in Washington DC, and a lot of people in various parts of the government that want to do that." "They want to say, we won't let these models trained in China be used in the US. There's a lot of people trying to do that now, and that would make us all worse off. I'm pretty nervous about that." Quinn Slack Amp

MTS

58,660 次观看 • 2 个月前

REPORT: Secret Chinese tech has been found inside U.S. solar equipment—and U.S. investigators say it could shut down the entire power grid. For the past nine months, these investigators have uncovered rogue communication devices buried in solar inverters and batteries made in China. These hidden parts were NOT listed in the manuals. But once investigators tore them open, they discovered a secret way to bypass security and take control remotely. If triggered, these backdoors could shut down power systems, damage the grid, and cause BLACKOUTS across the country, officials warned. Rep. August Pfluger said the CCP “stops at nothing” to break into our critical systems. “It’s about time we ramp up our efforts to show China that compromising us will no longer be acceptable,” he added. The threat is so serious, two senators introduced a bill to block these components. But it’s still sitting in committee—going nowhere. Because when the lights go out, it won’t just be your power—your country’s defenses go out with it. And you can understand how scary that is. Watch the full breakdown with .

REPORT: Secret Chinese tech has been found inside U.S. solar equipment—and U.S. investigators say it could shut down the entire power grid. For the past nine months, these investigators have uncovered rogue communication devices buried in solar inverters and batteries made in China. These hidden parts were NOT listed in the manuals. But once investigators tore them open, they discovered a secret way to bypass security and take control remotely. If triggered, these backdoors could shut down power systems, damage the grid, and cause BLACKOUTS across the country, officials warned. Rep. August Pfluger said the CCP “stops at nothing” to break into our critical systems. “It’s about time we ramp up our efforts to show China that compromising us will no longer be acceptable,” he added. The threat is so serious, two senators introduced a bill to block these components. But it’s still sitting in committee—going nowhere. Because when the lights go out, it won’t just be your power—your country’s defenses go out with it. And you can understand how scary that is. Watch the full breakdown with .

The Vigilant Fox 🦊

192,758 次观看 • 1 年前

🚨 Brandon Judd says “It’s Going to be Bad” when we learn why there are so many Illegal Chinese Military Aged Men in the US • Q: How do Chinese get out of China? —— A: They don’t have the money to get out of China… The cartels fund and smuggle them out of China. —— they pay the cartel back through slave labor, sex work, etc • Q: What about the military aged Chinese men? —— A: We don’t know what their purpose is here in the United States, but it’s going to be bad • Q: Do you believe the Chinese military aged men are being sent here by China? —— A: We believe so, yes. “Whether they’re spying on us or laying in wait… for some other purpose, we just don’t know.” Q: Does anyone monitor them or know where they are? —— A: No… when these people were released, we lose complete contact with them, unless they show up to their court appearances, which the majority of them do not.

🚨 Brandon Judd says “It’s Going to be Bad” when we learn why there are so many Illegal Chinese Military Aged Men in the US • Q: How do Chinese get out of China? —— A: They don’t have the money to get out of China… The cartels fund and smuggle them out of China. —— they pay the cartel back through slave labor, sex work, etc • Q: What about the military aged Chinese men? —— A: We don’t know what their purpose is here in the United States, but it’s going to be bad • Q: Do you believe the Chinese military aged men are being sent here by China? —— A: We believe so, yes. “Whether they’re spying on us or laying in wait… for some other purpose, we just don’t know.” Q: Does anyone monitor them or know where they are? —— A: No… when these people were released, we lose complete contact with them, unless they show up to their court appearances, which the majority of them do not.

MJTruthUltra

2,385,555 次观看 • 1 年前

Bloomberg warns that China’s AI price war may make profitability difficult for years. They should be more worried about Wall Street. If the price war continues, the real casualty may not be AI. It may be the entire valuation structure built around it. OpenAI and Anthropic are valued as though frontier intelligence will remain scarce, expensive, and capable of producing software-like margins. Hyperscalers are spending hundreds of billions on infrastructure based on that assumption. Nvidia’s valuation depends on those capital expenditures continuing for years. But Chinese models are driving token prices toward commodity levels. Usage can explode while revenue per token collapses. The servers become busier. The models become cheaper. The profits fail to appear. Nvidia has real revenue and real margins, so it is not the weakest link. But even Nvidia is priced on the belief that today’s extraordinary AI capital expenditure will remain economically justified. If cheaper models, better efficiency, and open weights destroy the expected returns on that infrastructure, Wall Street will not merely reprice OpenAI and Anthropic. It will reprice the entire AI chain. America built the AI bubble on Nvidia. Nvidia built its empire on TSMC. Taiwan built its economic security around TSMC. If Chinese efficiency destroys the economics of brute-force AI, the repricing will not stop in Silicon Valley. It will cross the Pacific and land in Hsinchu. AI may continue transforming the world while the AI bubble collapses. The internet survived 2000. The valuations did not.

Bloomberg warns that China’s AI price war may make profitability difficult for years. They should be more worried about Wall Street. If the price war continues, the real casualty may not be AI. It may be the entire valuation structure built around it. OpenAI and Anthropic are valued as though frontier intelligence will remain scarce, expensive, and capable of producing software-like margins. Hyperscalers are spending hundreds of billions on infrastructure based on that assumption. Nvidia’s valuation depends on those capital expenditures continuing for years. But Chinese models are driving token prices toward commodity levels. Usage can explode while revenue per token collapses. The servers become busier. The models become cheaper. The profits fail to appear. Nvidia has real revenue and real margins, so it is not the weakest link. But even Nvidia is priced on the belief that today’s extraordinary AI capital expenditure will remain economically justified. If cheaper models, better efficiency, and open weights destroy the expected returns on that infrastructure, Wall Street will not merely reprice OpenAI and Anthropic. It will reprice the entire AI chain. America built the AI bubble on Nvidia. Nvidia built its empire on TSMC. Taiwan built its economic security around TSMC. If Chinese efficiency destroys the economics of brute-force AI, the repricing will not stop in Silicon Valley. It will cross the Pacific and land in Hsinchu. AI may continue transforming the world while the AI bubble collapses. The internet survived 2000. The valuations did not.

𝘊𝘰𝘳𝘳𝘪𝘯𝘦

13,962 次观看 • 1 个月前

🚨 JUST IN: President Trump goes full PRO-DATA CENTER MODE, saying China will LOVE it if we stop building them This comes after X BUSTED a major Chinese bot influence operation against data centers "China could not be happier with this anti Data Center movement. Actually, they can’t believe it is happening!" "The only reason that communities throughout the U.S.A. should not want Data Centers is if they want to end up being backwards and poor. If they want to be successful and rich, with far lower taxes and jobs all over the place, let Data Reign." "The good news is that there are plenty of other places that want them. If we kill the Golden Goose, you will only have yourselves to blame." He's right. We have to win. And people's taxes can be SLASHED on top of it

🚨 JUST IN: President Trump goes full PRO-DATA CENTER MODE, saying China will LOVE it if we stop building them This comes after X BUSTED a major Chinese bot influence operation against data centers "China could not be happier with this anti Data Center movement. Actually, they can’t believe it is happening!" "The only reason that communities throughout the U.S.A. should not want Data Centers is if they want to end up being backwards and poor. If they want to be successful and rich, with far lower taxes and jobs all over the place, let Data Reign." "The good news is that there are plenty of other places that want them. If we kill the Golden Goose, you will only have yourselves to blame." He's right. We have to win. And people's taxes can be SLASHED on top of it

Eric Daugherty

133,187 次观看 • 25 天前

Elon Musk explains how data centers in Space could help the US compete with China in AI “Given that Chinese AI companies are doing as well as they are with a relatively small amount of compute, it seems that if they had a lot of compute, there’s a good chance they would be the leaders. The constraints on AI are essentially electricity and AI chips. China is very strong in digital AI, and it has far more electricity than the United States. In fact, China already has more electricity than the United States, Europe and India combined. The US government can prevent US companies from using Chinese models, but it can’t stop the rest of the world from using them. That won’t stop China from becoming the leader in AI. Digital intelligence comes down to how much AI compute, how many chips and how much electricity you have. The current constraint is electricity outside China. In China, the constraint is chips. China is also closer than most people realize to solving the lithography problem, which would enable it to make AI chips in very large volumes. Once we address the power constraint with AI data centers in space, the constraint will once again be chips outside China.”

Elon Musk explains how data centers in Space could help the US compete with China in AI “Given that Chinese AI companies are doing as well as they are with a relatively small amount of compute, it seems that if they had a lot of compute, there’s a good chance they would be the leaders. The constraints on AI are essentially electricity and AI chips. China is very strong in digital AI, and it has far more electricity than the United States. In fact, China already has more electricity than the United States, Europe and India combined. The US government can prevent US companies from using Chinese models, but it can’t stop the rest of the world from using them. That won’t stop China from becoming the leader in AI. Digital intelligence comes down to how much AI compute, how many chips and how much electricity you have. The current constraint is electricity outside China. In China, the constraint is chips. China is also closer than most people realize to solving the lithography problem, which would enable it to make AI chips in very large volumes. Once we address the power constraint with AI data centers in space, the constraint will once again be chips outside China.”

DogeDesigner

246,547 次观看 • 2 个月前

Turkish Deputy FM Levent Gümrükçü: When you remember how we ended up by buying the S-400, I'm going to put it into context there as well. The whole process of us acquiring an air defense system started in 2010. We opened many tenders—I think three or four tenders were opened in the first seven years or so—and every time the U.S. offer turned out to be the most expensive. That's understandable because Patriot systems are the best on the market right now. But they also turned out to be the least cooperative in terms of offsets, possibilities for technology sharing, or co-production of any of the components. More importantly, they turned out to be the least reliable as well, because every time we discussed with our U.S. colleagues whether they could guarantee the delivery of these urgently needed air defense systems, they said they were a democratic country with congressional oversight and there was no way they could guarantee delivery within a certain period of time. They also said that Türkiye's image in Congress was not good and that we had to work not only with the administration but with Congress as well.

Turkish Deputy FM Levent Gümrükçü: When you remember how we ended up by buying the S-400, I'm going to put it into context there as well. The whole process of us acquiring an air defense system started in 2010. We opened many tenders—I think three or four tenders were opened in the first seven years or so—and every time the U.S. offer turned out to be the most expensive. That's understandable because Patriot systems are the best on the market right now. But they also turned out to be the least cooperative in terms of offsets, possibilities for technology sharing, or co-production of any of the components. More importantly, they turned out to be the least reliable as well, because every time we discussed with our U.S. colleagues whether they could guarantee the delivery of these urgently needed air defense systems, they said they were a democratic country with congressional oversight and there was no way they could guarantee delivery within a certain period of time. They also said that Türkiye's image in Congress was not good and that we had to work not only with the administration but with Congress as well.

Clash Report

87,340 次观看 • 2 个月前

Scott Bessent (2023) says China won’t be able to boost productivity with AI “I was on a Zoom with one of the most prominent tech overlords. He did a trip around the world and he said, China will never let AI meaningfully into the private sector because they will be convinced that maybe Jack Ma with AI could take down the CCP.” “So you will not get this private sector productivity enhancement in China. He was of the view that Europe was already starting to regulate it away. And unfortunately the UK seems to be following European model.” “The US, as we tend to do in the US, he said was the wild, wild West. There could be incredible productivity gains for the US if we don't blow ourselves up."

Scott Bessent (2023) says China won’t be able to boost productivity with AI “I was on a Zoom with one of the most prominent tech overlords. He did a trip around the world and he said, China will never let AI meaningfully into the private sector because they will be convinced that maybe Jack Ma with AI could take down the CCP.” “So you will not get this private sector productivity enhancement in China. He was of the view that Europe was already starting to regulate it away. And unfortunately the UK seems to be following European model.” “The US, as we tend to do in the US, he said was the wild, wild West. There could be incredible productivity gains for the US if we don't blow ourselves up."

dnap

96,116 次观看 • 1 个月前

Anthropic admitted they built an AI so capable they were scared to release it and the number that explains why is 250. Anthropic's CFO Krishna Rao described in this clip what happened when they ran Mythos against an open source codebase that a previous frontier model had already analyzed. The prior model found 22 security vulnerabilities, Mythos found 250. In the same codebase, that the previous model had already reviewed and flagged as relatively clean. That number, more than 11 times as many vulnerabilities discovered is not just a benchmark improvement, it is a signal that there is an entire layer of software infrastructure that humanity has been operating under the assumption was secure and that assumption may no longer hold. The UK AI Security Institute independently evaluated Mythos Preview and confirmed what the internal numbers suggested. On expert level capture the flag challenges that no model could complete before April 2025, Mythos succeeded 73% of the time and it became the first model ever to complete a complex end-to-end attack range from start to finish, autonomously, without human guidance. The World Economic Forum called this a new security-driven era for AI, the Governor of the Bank of England publicly warned that Anthropic may have found a way to unlock the entire cyber-risk landscape, and the European Central Bank began quietly contacting financial institutions to assess their security posture. The response from Anthropic is what makes this story genuinely important. Rather than shelving the model or publishing it as a standard API release, Rao described a phased approach restricting access to a controlled group, focusing specifically on how the cyber capabilities can be used defensively rather than offensively and treating that framework as a template for how to release powerful but dangerous models in the future. The broader context makes that framing even more significant. AI generated code is already creating ten times more security vulnerabilities than human-written code, 63% of organizations reported experiencing an AI driven cyberattack in the past 12 months, and traditional signature-based security tools were built for a threat model that no longer describes the attack surface companies are defending against. Mythos represents a genuine leap in what autonomous security reasoning can do and it cuts both ways. The model that can find 250 vulnerabilities in a codebase a prior model rated as mostly clean is also, in the wrong hands, the model that can exploit those 250 vulnerabilities before a human defender has even finished reading the report. Anthropic's phased release strategy is not just a legal or PR decision, it is the most honest signal yet from a frontier lab that safety governance and capability development can no longer be treated as separate workstreams. The question is not whether this technology gets deployed, it is whether the institutions using it defensively stay ahead of the ones who will eventually use it offensively and whether the labs building it can keep those two timelines from inverting.

Anthropic admitted they built an AI so capable they were scared to release it and the number that explains why is 250. Anthropic's CFO Krishna Rao described in this clip what happened when they ran Mythos against an open source codebase that a previous frontier model had already analyzed. The prior model found 22 security vulnerabilities, Mythos found 250. In the same codebase, that the previous model had already reviewed and flagged as relatively clean. That number, more than 11 times as many vulnerabilities discovered is not just a benchmark improvement, it is a signal that there is an entire layer of software infrastructure that humanity has been operating under the assumption was secure and that assumption may no longer hold. The UK AI Security Institute independently evaluated Mythos Preview and confirmed what the internal numbers suggested. On expert level capture the flag challenges that no model could complete before April 2025, Mythos succeeded 73% of the time and it became the first model ever to complete a complex end-to-end attack range from start to finish, autonomously, without human guidance. The World Economic Forum called this a new security-driven era for AI, the Governor of the Bank of England publicly warned that Anthropic may have found a way to unlock the entire cyber-risk landscape, and the European Central Bank began quietly contacting financial institutions to assess their security posture. The response from Anthropic is what makes this story genuinely important. Rather than shelving the model or publishing it as a standard API release, Rao described a phased approach restricting access to a controlled group, focusing specifically on how the cyber capabilities can be used defensively rather than offensively and treating that framework as a template for how to release powerful but dangerous models in the future. The broader context makes that framing even more significant. AI generated code is already creating ten times more security vulnerabilities than human-written code, 63% of organizations reported experiencing an AI driven cyberattack in the past 12 months, and traditional signature-based security tools were built for a threat model that no longer describes the attack surface companies are defending against. Mythos represents a genuine leap in what autonomous security reasoning can do and it cuts both ways. The model that can find 250 vulnerabilities in a codebase a prior model rated as mostly clean is also, in the wrong hands, the model that can exploit those 250 vulnerabilities before a human defender has even finished reading the report. Anthropic's phased release strategy is not just a legal or PR decision, it is the most honest signal yet from a frontier lab that safety governance and capability development can no longer be treated as separate workstreams. The question is not whether this technology gets deployed, it is whether the institutions using it defensively stay ahead of the ones who will eventually use it offensively and whether the labs building it can keep those two timelines from inverting.

Milk Road AI

24,356 次观看 • 4 个月前

🚨Could Iran shut down the Strait of Hormuz if the US🇺🇸 and Israel attack? ‘What we worry about in the Middle East is the wider conflict or the reignition of the war between Israel and Iran, which would bring into question the role of the Strait of Hormuz, through which one third of all gas and one third of all oil used in the world goes through. So if there would be a major crisis or war between Israel and Iran, the Strait of Hormuz could be closed, and we would have an energy crisis of epic proportions. If push would come to shove, I’m sure they would do it. It would hurt them also. But if the very existence of Iranian leadership, or Iran as a state, would come into question, I’m sure they would do it. But I think that President Trump does not want to go there. So there is a backstop here, I think, in the US.’ -Tuomas Malinen, CEO of GnS Economics on Going Underground

🚨Could Iran shut down the Strait of Hormuz if the US🇺🇸 and Israel attack? ‘What we worry about in the Middle East is the wider conflict or the reignition of the war between Israel and Iran, which would bring into question the role of the Strait of Hormuz, through which one third of all gas and one third of all oil used in the world goes through. So if there would be a major crisis or war between Israel and Iran, the Strait of Hormuz could be closed, and we would have an energy crisis of epic proportions. If push would come to shove, I’m sure they would do it. It would hurt them also. But if the very existence of Iranian leadership, or Iran as a state, would come into question, I’m sure they would do it. But I think that President Trump does not want to go there. So there is a backstop here, I think, in the US.’ -Tuomas Malinen, CEO of GnS Economics on Going Underground

Going Underground

23,165 次观看 • 7 个月前

Kimi K3's weights are free to download. Almost nobody on earth can actually run them. 2.8 trillion parameters. The download is the easy part. Open weights used to mean you could self-host. At this scale it means you can read the file. So people download it and call the API anyway. Daily revenue up at least 6x since launch. Openness stopped being the opposite of a business model. It became the top of the funnel. The license says the quiet part. Past $20M in revenue you negotiate a contract. Past 100M users you display their name. Free for the developer. A contract for anyone the model actually earns from. And publishing the weights kills the argument before it starts. Nobody can claim the benchmarks were cooked. Independent indexes put K3 top three overall and first in frontend coding, with no one taking Moonshot's word for it. That's the whole trade. Give away the artifact, keep the infrastructure, get the audit for free. Check the last open model you praised. Ask whether you could serve it, or only download it.

Kimi K3's weights are free to download. Almost nobody on earth can actually run them. 2.8 trillion parameters. The download is the easy part. Open weights used to mean you could self-host. At this scale it means you can read the file. So people download it and call the API anyway. Daily revenue up at least 6x since launch. Openness stopped being the opposite of a business model. It became the top of the funnel. The license says the quiet part. Past $20M in revenue you negotiate a contract. Past 100M users you display their name. Free for the developer. A contract for anyone the model actually earns from. And publishing the weights kills the argument before it starts. Nobody can claim the benchmarks were cooked. Independent indexes put K3 top three overall and first in frontend coding, with no one taking Moonshot's word for it. That's the whole trade. Give away the artifact, keep the infrastructure, get the audit for free. Check the last open model you praised. Ask whether you could serve it, or only download it.

wast3

62,507 次观看 • 1 个月前

BIG | 🇨🇳🇹🇼🇷🇺🇪🇺‼️‼️ NATO Secretary General Mark Rutte is warning that a future crisis over Taiwan could rapidly become a two-theatre confrontation, with China potentially seeking Russian pressure against NATO in Europe while Beijing moves in the Indo-Pacific. Rutte said Western policymakers increasingly recognize that if China moved against Taiwan or elsewhere along the first island chain, Xi Jinping could turn to Vladimir Putin and ask Russia to “keep us busy” in Europe. Such a scenario would confront the United States and its allies with simultaneous security crises in Europe and the Indo-Pacific, stretching military resources across two strategically connected theatres. The warning does not mean NATO has intelligence showing that coordinated Chinese and Russian attacks are imminent. Rutte is describing a strategic scenario that he believes Western governments must prepare for as Russia rebuilds its military capacity and China increases pressure around Taiwan. His broader message is that European deterrence and Indo-Pacific security can no longer be planned separately, because a crisis in one region could be deliberately exploited in the other. See the latest updates with us: Visioner

BIG | 🇨🇳🇹🇼🇷🇺🇪🇺‼️‼️ NATO Secretary General Mark Rutte is warning that a future crisis over Taiwan could rapidly become a two-theatre confrontation, with China potentially seeking Russian pressure against NATO in Europe while Beijing moves in the Indo-Pacific. Rutte said Western policymakers increasingly recognize that if China moved against Taiwan or elsewhere along the first island chain, Xi Jinping could turn to Vladimir Putin and ask Russia to “keep us busy” in Europe. Such a scenario would confront the United States and its allies with simultaneous security crises in Europe and the Indo-Pacific, stretching military resources across two strategically connected theatres. The warning does not mean NATO has intelligence showing that coordinated Chinese and Russian attacks are imminent. Rutte is describing a strategic scenario that he believes Western governments must prepare for as Russia rebuilds its military capacity and China increases pressure around Taiwan. His broader message is that European deterrence and Indo-Pacific security can no longer be planned separately, because a crisis in one region could be deliberately exploited in the other. See the latest updates with us: Visioner

Visioner

146,845 次观看 • 7 天前

Sriram Krishnan on why Chinese models can distill off American models but American startups can't: "First of all, every model we have today is distilled off of all human knowledge. If you go back to the original GPT or the original Claude, they all had to derive off crawling the internet, crawling off all of our blogs and tweets and content. They kind of consume human knowledge to bootstrap that." "I'm going to steal this from Dean Meyer of Sequoia, who had a fantastic post yesterday. The situation which is bad today is that some of these models from other countries can train off American models, whereas if you are an American open weight model, it may be really confusing or challenging on whether you can distill off of other American models." "We kind of have a really uneven ecosystem here, where if you're a Chinese model, you could probably get a bunch of reasoning traces. But if you are a new Valley startup and you want to use some reasoning traces, you don't know what the legal situation is." "If you look at any American open source model today, they are using Chinese models as a teacher, or in a way as a part of the fine-tuning process. How do we make sure that is protected and enshrined, so if you're an American model company, you have the same level playing field as the Chinese models?" Sriram Krishnan

Sriram Krishnan on why Chinese models can distill off American models but American startups can't: "First of all, every model we have today is distilled off of all human knowledge. If you go back to the original GPT or the original Claude, they all had to derive off crawling the internet, crawling off all of our blogs and tweets and content. They kind of consume human knowledge to bootstrap that." "I'm going to steal this from Dean Meyer of Sequoia, who had a fantastic post yesterday. The situation which is bad today is that some of these models from other countries can train off American models, whereas if you are an American open weight model, it may be really confusing or challenging on whether you can distill off of other American models." "We kind of have a really uneven ecosystem here, where if you're a Chinese model, you could probably get a bunch of reasoning traces. But if you are a new Valley startup and you want to use some reasoning traces, you don't know what the legal situation is." "If you look at any American open source model today, they are using Chinese models as a teacher, or in a way as a part of the fine-tuning process. How do we make sure that is protected and enshrined, so if you're an American model company, you have the same level playing field as the Chinese models?" Sriram Krishnan

MTS

71,286 次观看 • 2 个月前