(From what I read) Early on, the AI labs were considering safety, when Altman/OpenAI went Silicon Valley “release & damn the icebergs” (and there _are_ icebergs).

I don’t scream loudly at Anthropic for defaulting to safety (cause I don’t really want to encourage the opposite).

Elon Musks timeline over the past 2 weeks is doubly insane when you consider the upcoming SpaceX IPO and how fundamentally, irrevocably tied he is to it's future.

It's like we just deciding that some things don't matter.

I’m not usually squeamish about tv - but Mehdir Hasan going on the PBDs Podcast was brutal.

Props to PBD for posting it.

It will be interesting to see if he still uses some of those talking points (or if he learns from the beating).

https://www.youtube.com/live/DT3gimXBzAI?si=T6CbKOtu_DZOE8lx

Why are our @ThinkstCanary booths always staffed by our engineers?

We get to do this:

“Someone came past yesterday to chat http://tokens.org... our emails didn't include the event name for aws api key triggers. He came back .. today…fix was already live”

I’ve mentioned this before: this is one of the oncoming trains for corp-security. We’ve long failed at least-privilege, but weren’t often punished for it.

Helen in HR (or Bob in accounts) didn’t know what to do with the extra perms they didn’t know they had.

Their agents will.