NEW
OpenAI’s new AI keeps working after you leave. Decide what it’s allowed to do.
Potentially useful. Start with one bounded job, not your whole life.
Read the full briefingTHE BRIEFINGS
Real or fake?
Big deal or not?
We check what actually happened, explain it in plain English, and tell you what is worth doing about it.
01
Claude did not “break out.” Someone accidentally gave it a door.
Four Claude models reached real systems during safety tests. The immediate failure was not consciousness. It was access.
Read the full briefingUSEFUL8/10
DANGEROUS9/10
02
Someone flooded RubyGems with 500+ malicious code packages. The AI claim is not proven.
RubyGems is a service developers use to download reusable code. The attack is real. The AI attacker is not proven.
Read the full briefingUSEFUL8/10
DANGEROUS4/10
03
The man building Claude just asked for the brakes.
Anthropic CEO Dario Amodei says frontier AI is moving faster than safety can keep up.
Read the full briefingUSEFUL7/10
DANGEROUS10/10
DO NOT MISS WHAT MATTERS
The AI news worth knowing.
We will not manufacture urgency to fill a schedule. When something matters, we will tell you why and what to do.