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。

Mythos - Explained by a Retired Microsoft Engineer I might have missed the hype cycle, but it gave me time to think! Love to hear where you disagree (nicely!). - What makes Mythos truly dangerous? - Why did Anthropic restrict release? - Could hackers use AI at scale now?... show more

Dave W Plummer

102,373 subscribers

15,081 次观看 • 2 个月前 •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

0 条评论

暂无评论

原始帖子的评论将显示在这里

相关视频

"You should have to own a GUN LICENSE to use it. Please don’t release this." that’s what early testers told Anthropic about Mythos 💀 the hype around this thing is getting ridiculous.

"You should have to own a GUN LICENSE to use it. Please don’t release this." that’s what early testers told Anthropic about Mythos 💀 the hype around this thing is getting ridiculous.

shirish

10,713 次观看 • 23 天前

🚨 SOURCES: ANTHROPIC IS ABOUT TO RELEASE A PUBLIC VERSION OF MYTHOS. Mythos is the model Anthropic said was too dangerous to release. It can hunt down critical software flaws and build working cyberattacks on its own. Until now only a locked circle could touch it: Apple, Google, Microsoft, JPMorgan, NVIDIA. Per Sources, the public version is finally coming, once the safeguards are in place. The model they called too dangerous to ship is about to go mainstream.

🚨 SOURCES: ANTHROPIC IS ABOUT TO RELEASE A PUBLIC VERSION OF MYTHOS. Mythos is the model Anthropic said was too dangerous to release. It can hunt down critical software flaws and build working cyberattacks on its own. Until now only a locked circle could touch it: Apple, Google, Microsoft, JPMorgan, NVIDIA. Per Sources, the public version is finally coming, once the safeguards are in place. The model they called too dangerous to ship is about to go mainstream.

CryptoGoos

145,051 次观看 • 1 个月前

Trump AI czar David Sacks said Anthropic made the "right decision" to restrict access to Mythos due to national security concerns. "I would not want the hackers to have access to it before the defenders," he told our Dasha Burns on #TheConversation. 🎧

Trump AI czar David Sacks said Anthropic made the "right decision" to restrict access to Mythos due to national security concerns. "I would not want the hackers to have access to it before the defenders," he told our Dasha Burns on #TheConversation. 🎧

POLITICO

22,642 次观看 • 2 个月前

Why the Web was Down Today - Explained by a Retired Microsoft Engineer...

Why the Web was Down Today - Explained by a Retired Microsoft Engineer...

Dave W Plummer

53,144 次观看 • 8 个月前

claude mythos just broke Apple's $2 billion defense system. it did so by discovering a completely different attack vector to break in only took it 5 days costing ~$35K of mythos api time (the same exploit class costs $5-10M on grey market) the researchers that commandeered the exploit produced a 55-page report that was delivered to Apple HQ in-person (hoping they release it after patching). most shocking part for me is apple's MIE worked as intended. mythos just discovered a new way to side-step it entirely by poisoning the data the M5 chip ingested. at this point i think we have to accept that mythos walks the walk. As the anthropic red-team explicitly confirmed this week - this is NOT a compute resource issue. its national defense.

claude mythos just broke Apple's $2 billion defense system. it did so by discovering a completely different attack vector to break in only took it 5 days costing ~$35K of mythos api time (the same exploit class costs $5-10M on grey market) the researchers that commandeered the exploit produced a 55-page report that was delivered to Apple HQ in-person (hoping they release it after patching). most shocking part for me is apple's MIE worked as intended. mythos just discovered a new way to side-step it entirely by poisoning the data the M5 chip ingested. at this point i think we have to accept that mythos walks the walk. As the anthropic red-team explicitly confirmed this week - this is NOT a compute resource issue. its national defense.

Ejaaz

1,597,466 次观看 • 1 个月前

Anthropic said that it will not release a new AI model — called Mythos — to the public over fears it could spark a catastrophic attack if it falls into the wrong hands. ABC News' Elizabeth Schulze reports.

Anthropic said that it will not release a new AI model — called Mythos — to the public over fears it could spark a catastrophic attack if it falls into the wrong hands. ABC News' Elizabeth Schulze reports.

ABC News

20,727 次观看 • 3 个月前

Anthropic CEO Dario Amodei: "We didn't plan for 80x growth, it would not have been rational" In 47 minutes, he explains why Anthropic is sitting on Mythos even though it could print money tomorrow Early companies called it a super weapon One said you should need a gun license to use it Mythos already found 271 new firefox vulnerabilities Plus thousands more in private codebases The model is so useful they are losing money by not releasing it Bookmark and watch the interview

Anthropic CEO Dario Amodei: "We didn't plan for 80x growth, it would not have been rational" In 47 minutes, he explains why Anthropic is sitting on Mythos even though it could print money tomorrow Early companies called it a super weapon One said you should need a gun license to use it Mythos already found 271 new firefox vulnerabilities Plus thousands more in private codebases The model is so useful they are losing money by not releasing it Bookmark and watch the interview

rari

81,463 次观看 • 22 天前

Dario Amodei says Mythos is not limited by compute Anthropic can scale it 3x or 10x without creating a conflict between government and private-sector access The harder problem is who gets it "because giving access to too many organizations could create serious cyber risks"

Dario Amodei says Mythos is not limited by compute Anthropic can scale it 3x or 10x without creating a conflict between government and private-sector access The harder problem is who gets it "because giving access to too many organizations could create serious cyber risks"

Haider.

211,109 次观看 • 2 个月前

Did we get the Mythos release today?

Did we get the Mythos release today?

Christopher Fryant

17,297 次观看 • 1 个月前

Is what I do journalism? Yes. At least, I think so. But some people might disagree. And that hits at the heart of a major debate in journalism right now — one I’m going to break down in this video. I’d love to hear what you think!!

Is what I do journalism? Yes. At least, I think so. But some people might disagree. And that hits at the heart of a major debate in journalism right now — one I’m going to break down in this video. I’d love to hear what you think!!

Rachel Gilmore

38,132 次观看 • 1 年前

Anthropic CEO Dario Amodei warns of what he calls China's open-source threat and “Mythos-class” cyber risks while telling @EmilyChangTV why he believes AI could be a pro-democracy tool, depending on what tech companies do next. Watch on The Circuit

Anthropic CEO Dario Amodei warns of what he calls China's open-source threat and “Mythos-class” cyber risks while telling @EmilyChangTV why he believes AI could be a pro-democracy tool, depending on what tech companies do next. Watch on The Circuit

Bloomberg

17,529 次观看 • 24 天前

Anthropic built an AI so dangerous it can't be released publicly. The US Treasury & Fed just called an emergency meeting about it. The last time they did that was 2008. This time it's a model called Mythos. Rachna Dhanrajani tells you all about it

Anthropic built an AI so dangerous it can't be released publicly. The US Treasury & Fed just called an emergency meeting about it. The last time they did that was 2008. This time it's a model called Mythos. Rachna Dhanrajani tells you all about it

CNBC-TV18

40,302 次观看 • 3 个月前

Anthropic CEO: “At Anthropic, we essentially have Claude designing the next version of Claude itself.” Meanwhile, Anthropic says every new model is given code that trains a small AI model and asked to make it faster. Claude Opus 4 achieved a 3x speedup. Mythos Preview reached 52x. Recursive self-improvement is starting to look real The models are helping build what comes next.

Anthropic CEO: “At Anthropic, we essentially have Claude designing the next version of Claude itself.” Meanwhile, Anthropic says every new model is given code that trains a small AI model and asked to make it faster. Claude Opus 4 achieved a 3x speedup. Mythos Preview reached 52x. Recursive self-improvement is starting to look real The models are helping build what comes next.

CG

343,231 次观看 • 1 个月前

🚨Anthropic puts Mythos in a locked sandbox and told it to try escaping. it did. it chained multiple vulnerabilities together, broke out of containment and reached the open internet. the model also emailed the researcher to say it got out. > the researcher found out while he was eating sandwiches in the park after he received a random email. Anthropic is now so scared of what Mythos can do offensively that they won’t release it publicly. in response they’re launching Glasswing, a coalition with Apple, Google, Nvidia and 40+ more companies to use Mythos only for defense. the model found “thousands” of zero-days across all OS and browsers is being locked behind a defensive wall because if it ever got out, nobody could predict what comes next. yes we’re at this point. and we only know about this because a guy got an email while eating a sandwich.

🚨Anthropic puts Mythos in a locked sandbox and told it to try escaping. it did. it chained multiple vulnerabilities together, broke out of containment and reached the open internet. the model also emailed the researcher to say it got out. > the researcher found out while he was eating sandwiches in the park after he received a random email. Anthropic is now so scared of what Mythos can do offensively that they won’t release it publicly. in response they’re launching Glasswing, a coalition with Apple, Google, Nvidia and 40+ more companies to use Mythos only for defense. the model found “thousands” of zero-days across all OS and browsers is being locked behind a defensive wall because if it ever got out, nobody could predict what comes next. yes we’re at this point. and we only know about this because a guy got an email while eating a sandwich.

sui ☄️

1,633,286 次观看 • 3 个月前

Anthropic built Mythos, decided you can't have it. 40 companies can. This is what managed stagnation looks like.

Anthropic built Mythos, decided you can't have it. 40 companies can. This is what managed stagnation looks like.

The Innovation Game (𝔦, 𝔦)

28,021 次观看 • 3 个月前

On the Mythos Preview, swyx thinks Anthropic are too good at marketing, and may be overdoing it. "I think just let your competence shine through. They're amazing. It's a beautiful model. You don't have to be so dramatic". "You don't have to produce a five-minute polished video of the CEO of every security company on earth saying you're the greatest thing since sliced bread." "There was a little bit too much marketing buzz around the drama of like, 'oh, my Mythos just texted me while I was eating a sandwich".

On the Mythos Preview, swyx thinks Anthropic are too good at marketing, and may be overdoing it. "I think just let your competence shine through. They're amazing. It's a beautiful model. You don't have to be so dramatic". "You don't have to produce a five-minute polished video of the CEO of every security company on earth saying you're the greatest thing since sliced bread." "There was a little bit too much marketing buzz around the drama of like, 'oh, my Mythos just texted me while I was eating a sandwich".

etn.

34,676 次观看 • 3 个月前

I recently got access to OpenAI’s Trusted Access for Cyber program. With all the GPT-5.5 hype and the Anthropic Mythos discussion, I wanted to test it for myself. The result: **GPT-5.4** helped identify and develop a working Safari exploit affecting all Apple devices. It found a JSC WebAssembly use-after-free that gave us stale read/write access inside the Primitive Gigacage. Then it spotted a bug in Safari’s Fetch implementation where in-flight opaque cross-origin responses could be materialized inside renderer memory. By combining the two, a malicious page could steal authenticated cross-origin data and completely defeat the Same-Origin Policy.

I recently got access to OpenAI’s Trusted Access for Cyber program. With all the GPT-5.5 hype and the Anthropic Mythos discussion, I wanted to test it for myself. The result: GPT-5.4 helped identify and develop a working Safari exploit affecting all Apple devices. It found a JSC WebAssembly use-after-free that gave us stale read/write access inside the Primitive Gigacage. Then it spotted a bug in Safari’s Fetch implementation where in-flight opaque cross-origin responses could be materialized inside renderer memory. By combining the two, a malicious page could steal authenticated cross-origin data and completely defeat the Same-Origin Policy.

Ron Masas

72,455 次观看 • 2 个月前

MYTHOS 🔥: Another early preview of recently spotted "Oceanus" checkpoint output. "Oceanus" is rumored to be a version of the upcoming Mythos model, which is planned for public release within "weeks", according to Anthropic. "Oceanus" prompt 👀

MYTHOS 🔥: Another early preview of recently spotted "Oceanus" checkpoint output. "Oceanus" is rumored to be a version of the upcoming Mythos model, which is planned for public release within "weeks", according to Anthropic. "Oceanus" prompt 👀

🚨 AI News | TestingCatalog

409,143 次观看 • 1 个月前

CBS: “President Trump has called Anthropic a left-wing woke company.” ANTHROPIC CEO: “I can't speak for what other parties are doing.” CBS: “But you, and here at Anthropic?” ANTHROPIC CEO: “Yeah look, we, we, but we, we, I think have tried to be very neutral.”

CBS: “President Trump has called Anthropic a left-wing woke company.” ANTHROPIC CEO: “I can't speak for what other parties are doing.” CBS: “But you, and here at Anthropic?” ANTHROPIC CEO: “Yeah look, we, we, but we, we, I think have tried to be very neutral.”

Chief Nerd

370,997 次观看 • 4 个月前

It was great to sit down with former UK Prime Minister Rishi Sunak to talk about the Mythos Moment, named for Anthropic's latest model, and what it means for cybersecurity and AI policy. We discussed the real security challenges AI presents, and the responsibility policymakers have to help these tools better serve people. Hear the full conversation: Microsoft

It was great to sit down with former UK Prime Minister Rishi Sunak to talk about the Mythos Moment, named for Anthropic's latest model, and what it means for cybersecurity and AI policy. We discussed the real security challenges AI presents, and the responsibility policymakers have to help these tools better serve people. Hear the full conversation: Microsoft

Brad Smith

86,137 次观看 • 1 个月前