Anthropic's "too dangerous" AI was accessed by guessing the URL

[davidgerard@circumstances.run]

Anthropic's "too dangerous" AI was accessed by guessing the URL

l o l

https://boingboing.net/2026/04/23/anthropics-too-dangerous-ai-was-accessed-by-guessing-the-url.html

the story they cribbed:
https://www.theregister.com/2026/04/22/anthropic_mythos_hype_nothingburger/

[oxy@social.bsdlab.au]

@davidgerard can the crash come already the stupidity of these guys is getting beyond boring

[bms48@mastodon.social]

@davidgerard https://www.youtube.com/watch?v=znWxaDZ2OjY "One is that the LLM systems are designed in such a way that they cannot tell us anything about language, learning, or other aspects of cognition, a matter of principle, irremediable.... The reason is elementary: The systems work just as well with impossible languages that infants cannot acquire as with those they acquire quickly and virtually reflexively." -- Noam Chomsky

[godsoewilliam@ecoevo.social]

@davidgerard@circumstances.runif only they had some super intelligence that could have pointed out this cybersecurity flaw.

[perigee@rage.love]

@davidgerard It seems very (sarcasm) responsible and ethical to me to put a "too dangerous" AI straight up on the Internet. Are we vibe networking now?