Hacker News: InkCanon

New comment by InkCanon in "Our position on open-weights models"

InkCanon — Tue, 28 Jul 2026 04:26:53 +0000

This argument is old and goes back to cyber security. Things like government backdoors are the same. It's true centralizing power in the hands of a "good" would theoretically be good. But the key part is the "good". I think after repeated experiences with governments and corporations, no rational person would ever assume them to be good. There are simply too many incentives to be bad.

New comment by InkCanon in "Sarah's Wager"

InkCanon — Tue, 21 Jul 2026 15:35:32 +0000

I'm also at a loss. It feels like some rube goldberg machine of connections among elites, bringing up people with no apparent connection to the article other than being big and in some way connected to someone else in it.

New comment by InkCanon in "Sarah's Wager"

InkCanon — Tue, 21 Jul 2026 15:32:17 +0000

I can put up with a little bit of dramatic name dropping in an article at the start, but it never stops. Almost every paragraph is reads like some snappy autobiography of someone trying to impress the reader. I'm not even sure what the article is about, other than name dropping.

New comment by InkCanon in "Control the Ideas, Not the Code"

InkCanon — Mon, 13 Jul 2026 14:14:41 +0000

Not specifically about the comment you replied to, but there has been an absolute Cambrian explosion of plausible sounding projects that use really complicated words that are just AI psychosis. Recently there was someone who posted a huge Lean proof on GitHub supposedly of proving P=NP. The problem is AI is really good at deluding people if that person doesn't have domain specific knowledge. AI severely makes dunning Kruger worse

New comment by InkCanon in "A global workspace in language models"

InkCanon — Tue, 07 Jul 2026 11:13:27 +0000

They phrase it in a confusing way, with a lot of fluff. But to my understanding what happens is 1. At some part in the transformer, calculate the Jacobian (matrix gradients for that forward pass). Do this for 1000 prompts and average them. 2. For another pass take an "activation vector" (the input vector x to that part). Apply the matrix in step 1 to this vector. 3. Unembed the output to get a token. Presumably to make this have meaning, they use the reading of past tokens after embedding, instead of a random vector in the middle of the transformer. The high level idea appears to be you're trying to find which token, if the value were changed slightly, creates the largest downstream change.

New comment by InkCanon in "Anthropic CEO: Open-Source AI is getting dangerous (2023)"

InkCanon — Mon, 29 Jun 2026 11:12:28 +0000

Completely stopping people from using open source models is impossible, but if it became a federal law there could be incredible pressure exerted on US companies. For example, OpenRouter might be compelled to not accept open source requests on threat of fines.

New comment by InkCanon in "Anthropic CEO: Open-Source AI is getting dangerous (2023)"

InkCanon — Mon, 29 Jun 2026 09:50:11 +0000

Paraphrasing him, he says that open source software which tool tens or hundreds of millions to build should be classified as a different category. And the implication being they should be banned. Why?

New comment by InkCanon in "Anthropic says Alibaba illicitly extracted Claude AI model capabilities"

InkCanon — Fri, 26 Jun 2026 14:09:03 +0000

I feel obliged to point out the disingenuousness of what this post says and the post it quotes. The most egregious parts are payment fraud and reselling - these are speculated but not actually said to be known to be happening, which you have left out.

1. Claims of payment fraud. I actually clicked the BBC article linked about payments fraud it referred to. It was an article about a criminal syndicate stealing credit cards. It mentions buying cryptocurrencies, AI API purchases are not mentioned.

2. The claims that they are reselling the chats to AI labs. The post you cite is speculating it could be, but this is unverified.

The claims of reselling is also bizzare. Arbitrary user prompts are low quality data. If I were an AI lab, wouldn't I just pay for the API proxy and get targeted output for far less?

Also calling it bot accounts is a stretch. Bots mimick human input. These are proxy accounts.

New comment by InkCanon in "Anthropic says Alibaba illicitly extracted Claude AI model capabilities"

InkCanon — Fri, 26 Jun 2026 08:53:01 +0000

The openness of AI is currently being held up only by Chinese companies (previously Meta, but they stopped). They're not saints, but there is not even a question in the open weight/HF community that the immense mass of Chinese talent, knowledge and resources are the only thing stopping a monopoly/duopoly from forming. In a very Cathedral vs Bazaar-esque way, China severely lacks compute, but are extremely ingenious in coming up with new optimizations, architectures, etc, which they all detail in their papers.

New comment by InkCanon in "Anthropic says Alibaba illicitly extracted Claude AI model capabilities"

InkCanon — Fri, 26 Jun 2026 08:45:57 +0000

"You're trying to kidnap what I've rightfully stolen."

New comment by InkCanon in "Anthropic says Alibaba illicitly extracted Claude AI model capabilities"

InkCanon — Fri, 26 Jun 2026 08:43:24 +0000

I'd hazard anthropic perceives their number 1 enemy as open weight models. If alternatives to their business exist (which is mainly coding tokens currently), they will get into a nasty fight and the nightmare of all tech companies - losing their monopoly. It threatens their ability to extract value, and could reduce their valuation to a tenth of what it was. They cannot make open weight models worse, so they're using lawfare to try and get them banned. And of course we previously know they attempted to block Chinese companies under the guise of national security by lobbying for restricting GPU sales.

New comment by InkCanon in "Anthropic says Alibaba illicitly extracted Claude AI model capabilities"

InkCanon — Thu, 25 Jun 2026 09:15:22 +0000

Anthropic is master of Newspeak (see previously bugs -> vulnerabilities wrt Mythos). Distillation violates their terms of service, which is a civil offense, not a criminal one. It is not illicit, illegal nor breaks any laws.

New comment by InkCanon in "Project Fetch: Phase Two"

InkCanon — Sun, 21 Jun 2026 03:42:21 +0000

Is that a Unitree?

New comment by InkCanon in "Claude Corps"

InkCanon — Sat, 13 Jun 2026 12:59:13 +0000

I'm confused, is the pitch here to train (to the extent that using an LLM needs training) like Red Cross volunteers to use Claude?

New comment by InkCanon in "How Terry Tao became an evangelist for AI in math"

InkCanon — Fri, 12 Jun 2026 09:59:55 +0000

The checking has nothing to do with AI, despite the (massively funded) marketing done to make you think so. It is based on formal methods/theorem provers.

New comment by InkCanon in "“Too dangerous to release” or just too expensive?"

InkCanon — Fri, 15 May 2026 14:51:45 +0000

Also, slightly stretching the definition of terms consecutively, so the multiplicative meaning is really far from the truth. For example, 271 vulnerabilities were really mostly bugs - generally incorrect states, but which almost never led to any exploit.

New comment by InkCanon in "“Why not just use Lean?”"

InkCanon — Tue, 28 Apr 2026 12:26:32 +0000

Wait till you hear what Rocq was originally called!

New comment by InkCanon in "The West forgot how to make things, now it’s forgetting how to code"

InkCanon — Sun, 26 Apr 2026 10:16:51 +0000

The Stinger is an anti air weapon, the Javelin is an anti tank weapon.

New comment by InkCanon in "There Will Be a Scientific Theory of Deep Learning"

InkCanon — Sat, 25 Apr 2026 12:20:24 +0000

The field is massively hampered by the wishful mnemonics and anthropomorphization of LLMs. For example, even the hallucination idea arbitrarily assigns human semantics to LLM results. By the actual mathematical principles by which LLMs work, any hallucination is another output, with no clear definition between it and every other output.

New comment by InkCanon in "Qwen3.6-Max-Preview: Smarter, Sharper, Still Evolving"

InkCanon — Tue, 21 Apr 2026 06:26:46 +0000

The more accurate version is only Chinese companies (plus Facebook briefly) really open source their frontier models. The rest are non frontier. They are either older or specialized for something.