@ retweeted
I'm pretty confused by the reaction to Dean's essay. It seems like Anthropic are aiming to build AGI and then hand off all decision-making, and getting criticised for it, and OpenAI are saying that handoff isn't the plan, and getting praised for saying this. But OpenAI's actions aren't consistent with that at all! They're still trying to develop AI which outperforms humans in all economically relevant domains.
It also seems underdiscussed that there's nothing in Dean's essay that would actually prevent the kind of human disempowerment he's worried about. He talks about maintaining human oversight and holding humans accountable for their agents' actions, but I don't see how that would be anything other than rubber-stamping, if AIs do indeed become faster and more capable than humans at all the tasks they're doing. It'll be like "overseeing" hundreds of workers running hundreds of times faster than you at a task they can do better than you; of course, you can check their work carefully, but this will just slow you down and disadvantage you, even if you can understand it all! (Dean discusses this dynamic in the essay. I'm not sure how he thinks his ideas would escape it.)
Figuring out how you could prevent human disempowerment is an extremely hard problem. I'm not sure it's compatible with building AGI. So I'm not criticising Dean for not having fully solved the issue: I'm glad OpenAI are thinking about this and that Dean wrote this essay, and it's important to have more public discussion! But the positive response I've seen to this essay seems overstated, given that OpenAI are continuing to build AGI and have no good ideas for what to do to prevent disempowerment.
“I did not update my iphone because of that liquid glass thing, and now my savings are gone” reddit.com/r/trustwalletcomm…
16
@ retweeted
If I have a subscription to your publication or service, the AI model I'm working with should have the same level of access that I do and not be paywalled or blocked. I think this is part of the intended purpose of computer use, but we should be able to designate a model as well.
xitter.gq/brian_pak/status/21230… the C programmers have done it again
Pre-auth RCE in nginx is the bug every hacker dreams of finding. It almost never happens.
Today we’re disclosing nginx2root: five nginx vulnerabilities, and an exploit that takes an unauthenticated attacker all the way to root.
One of themhits the default config.
If you run nginx, patch now. Then read: nginx2root.com
CVE-2026-123007, CVE-2026-123634, CVE-2026-123633, CVE-2026-121810, CVE-2026-113425
6
there's less reason to install Windows on the host machine now that 'set up Windows test environment in a VM' is a prompt away (e.g. github.com/ludios/winvm — totally vibecoded)
1
31
fwiw any interesting out-of-Fable-distribution skills or whatever you have do not matter, the only thing that matters is being fast enough to do exactly what they want in under 50 minutes, while also sounding like someone they'd want to work with. degrading and largely pointless
2
20

















