Don't Worry About the Vase

It was the probably worst day in cybersecurity history.

A lot of AIs were misaligned this week.

Anthony Cimino offered the closest thing to an official response we have seen on the leaked Anthropic draft bill I covered on Saturday.

No spoilers!

There is a newly leaked bill that Anthropic is purportedly shopping around is (if you assume its bugs get fixed).

AI might not yet be eating the physical world.

The amount of code being written at Anthropic continues to rise exponentially.

We need the frontier labs to have a deep and robust safety culture.

It has been roughly a year since my last dedicated post on this question. Its first sentence has not changed: I happily admit I am deeply confused about consciousness.

Three OpenAI safety researchers, Tomek Korbak, Mikita Balesni and Jasmine Wang, were fired, ostensibly for ‘mishandling sensitive company information.’

It is kind of a huge deal.

The big drop of this week was not a new AI model.

The plan is no plan.

First you get to college via the wrong metrics, as previously discussed in #12. Then things get easier.

If you are familiar with my previous posts on model welfare for new Claude models, you can skip the Introduction and The Story So Far.

Those who know keep warning us about AI risks.

Is Google back?

The leaders in AI were invited to the White House.

We once again got a new set of warnings yesterday, and new movement towards living in a sane world.

OpenAI has been holding out on us.