2,266 karma · joined May 10, 2013
Here's the macOS kernel-enforced sandbox profile that gets applied to processes spawned by the LLM: https://github.com/openai/codex/blob/main/codex-rs/core/src/...
I think skepticism is healthy here, but there's no need to just guess.
The point is that Codex can (by default) run commands on its own, without approval (e.g., running `make` on the project it's working on), but they're subject to the imposed OS sandbox.
This is controlled by the `--sandbox` and `--ask-for-approval` arguments to `codex`.
At least for Codex, the agent runs commands inside an OS-provided sandbox (Seatbelt on macOS, and other stuff on other platforms). It does not end up "making the agent mostly useless".
https://s3data.computerhistory.org/brochures/apple.applei.19...
This resolves any number of heads, but the resulting tree of the merge is always
that of the current branch head, effectively ignoring all changes from all other
branches. It is meant to be used to supersede old development history of side
branches. Note that this is different from the -Xours option to the ort merge strategy.You can hire the best, but coordination among a group of people scales quadratically.
The answer, unfortunately, is that features need to be sunsetted/removed, the engineering org shrunk, and for a smaller group to concentrate on a reliable product core.
(I still disagree with your broader point, since I don't think you can meaningfully draw a "has skin in the game" line.)
Maybe those transient homeowners are the ones who shouldn't get to vote...
Also, basically every pro and semi-pro sports stadium nowadays has cell-phone-contactable security that you can summon to handle situations like these. The threat of being kicked out of his $250 seats is way more of a threat than that of being confronted by a "mega-nerd".
I wouldn't make a habit of contacting security over every little annoyance, but if they're obnoxiously blocking an old lady, that's the time to use it.
P.S.: your karma is currently 1337, sweet
Personally, I use an 11-year-old machine and have had to add userscript hacks to certain major Web sites to work around bugs in CSS grid (not the "lanes" described here).
At least new JavaScript features can be "polyfilled" or whatever. Maybe sites could check for CSS feature support too? But they seem not to.
For example, the demo page linked in the article fails pretty unusably for me. All the images take up nearly the full viewport width.
I fear the author has missed the point of the "Project Vend" experiments, the original write-ups of which are available here (and are, IMO, pretty level-headed about the whole thing):
https://www.anthropic.com/research/project-vend-1
https://www.anthropic.com/research/project-vend-2
The former contains a section titled "Why did you have an LLM run a small business?" that attempts to explain the motivation behind the experiment.
And FreeBSD isn't actually POSIX-certified anyway!
The real consideration here is simply that there are tons of existing rc scripts for BSDs, and switching them all would be a large task.