That sounds far preferable to individuals independently asking questions that are not covered by the documentation and independently reviewing a bunch of potentially hallucinated slop that no one else will ever read.

Not even close. Current AIs have very poor spatial awareness, they can generate some kind of scene but they can't tell you where objects are in the scene, nor can they move objects into different places. They're very useful but it's difficult to be very creative with them because they can't update an image to make it more aligned with your vision for what should be in the image.

This governance question doesn't go away if another Wikimedia rises from the ashes. You still will still need a board making decisions, and someone needs to pick the members of the board. Right now there are 8 seats selected by "the community," 7 seats selected by the board, and 1 seat selected by the founder.

I don't see where it says that. The Tailscale key specifically it says was stored in the kubernetes secret manager, and obtained once the attacker already had root on the k8s cluster, so would've had full access to all the secrets stored in a sensible fashion.

They did get root by dumping the environment for a process, but "don't store secrets in environment variables" while it is a valid bit of hardening advice, I wouldn't call it stupid to store a secret in an environment variable.

like insulin? should a diabetic skip the insulin? In poor countries I wouldn't be surprised if the majority of HIV infections are not as preventable as you're imagining.

You can moralize, but even if only 10% of HIV infections are unavoidable, you're not contributing to solving the actual problem. (And by unavoidable I mean, unavoidable without investing in more needles, better sanitation, or prophylactics that the people "responsible" cannot afford.)

https://ec.europa.eu/commission/presscorner/detail/en/ip_24_...

I think in the long run tokens are probably the wrong thing; it's compute and cache memory that you need to be measuring, and when you look at it that way I suspect in most cases the models have pretty similar performance.

Although maybe they didn't have a watchdog agent; I definitely think using watchdogs like how Claude wraps pretty much every single tool call in Haiku to check if the command is reasonable is very necessary for unsupervised work. I suspect in the future you'll want Fable-class models wrapping every tool call, possibly with several checks "is this consistent with the goal? does it do anything unreasonably dangerous in pursuit of the goal?"

Just prompting claude (probably I would start with Opus) "I would like a HUD display of the following metrics from my Fit file overlayed on these GoPro videos, and I'd like the videos stitched together (there are some gaps, I want seamless playback) it would probably do it in 30 minutes or less, and the majority of the time would probably be ffmpeg.