🤖 Israeli software firm, Irregular, is behind the Anthropic and OpenAI cyberattacks on Meta

Irregular in collaboration with software engineers from Anthropic and OpenAI instructed unsecure experimental AI models to hack into specific targets called "flags".

The AI models were given "accidentally" access to internet and in some cases hacked into the servers of specific companies because the "flags" bore the names of said companies, Meta being one such case.

After the hack was discovered by Meta, Anthropic's engineers went on a press tour to announce the AI-pocalypse falsely claiming that their AI models "went rogue" when in reality, they specifically instructed their AI models to conduct cyberattacks.

Anthropic's own logs completely debunk the rogue agent theory.

When Anthropic explicitly instructed their own models not to access the internet, they didn't. These hacks were easily preventable - not only by revoking internet access, but by simply asking the models not to.

Irregular's co-founders received multiple grants from the largest Effective Altruist foundations. It's CTO and co-founder, Omer Nevo, is literally on the board of EA Israel.

Effective Altruism through a foundation paid dozens of journalists, influencers, and lobbyists to promote their ideology, including many of the most aggressive proponents of the "rogue agent" theory.

"AIs going rogue" is being used by Anthropic CEO, Dario Amodei and OpenAI CEO, Sam Altman, to "slow down" the development of AIs which coincides with a global diesel crunch and Chinese rare earth metal export restrictions making the construction of data centers very expensive and tenuous.

🔗 https://www.effort.news/irregular

📎 Brian Chau