Skip to content

Comment on AI Doomlord Jacob Coxon's Media Tour Has Begun

Comments

kg

The Anthropic employee quoted in the article is doing a great job of saying things that no PR/communications department would ever, ever want an employee to say:

Jacob is correct here—we really do earnestly believe AI could kill all humans! I personally think it is >10% within the next decade. I believe Anthropic is trying its best, but we do not yet have a plan to solve alignment for superintelligence and are not clearly on track to.

The goal isn’t positive PR. It’s to have governments take note and come down hard on Anthropic and OpenAI with regulations that they’ll ultimately have a hand in writing or providing testimony for.

Those regulations will ever so coincidentally make it difficult for anyone else to get as big as they are, and demonize open weights, Chinese AI, abliteration, etc

abliteration

OK, so I looked up this word, and TIL.

In the Infocom game, Hitchhiker's Guide to the Galaxy Arthur Dent struggles with two inventory items, which are named "no tea" and "tea". He (you) has suffered through the entire game carrying "no tea", but when you finally find real tea, the interpreter refuses to hold both at once, which becomes problematic.

So eventually you enter a maze, and discover that it is your own brain, wherein is contained a small black particle labelled "COMMON SENSE" and, it turns out that the object is to dislodge and discard that common sense from your brain.

Then, upon exiting back into normal space, you/Arthur can pick up tea and also "no tea" and hold both in your inventory at once, having successfully abliterated your guardrails against nonsense.

But this is more like an LLM holding "tea" and "no tea" at the same time, and you're trying to get it to drop one or the other and it refuses.

Which upstarts are the frontier labs threatened by except for each other? The "regulatory capture" arguments make zero sense. Especially since their lobbying arms are trying to reject any kind of regulation: https://x.com/AlexBores/status/2097846545408262403

Deepseek, Kimi, Qwen.

AboutSource Built by g1lg1l

Hackerly is an independent reader for Hacker News, built on the public HN API. Not affiliated with Y Combinator.