Hacker News: tomrod

New comment by tomrod in "AI is a tool and using it well takes skill"

tomrod — Thu, 15 Oct 2026 02:30:46 +0000

Oh have I got some stories.

New comment by tomrod in "A look at the math that OpenAI claims computes faster Fourier transforms"

tomrod — Thu, 15 Oct 2026 01:20:51 +0000

That's absolutely fascinating. I need to read up on this -- thanks for sharing!

New comment by tomrod in "OpenCode 2.0"

tomrod — Wed, 14 Oct 2026 22:55:10 +0000

Nice work!

Are there any recommended sandbox patterns to use? Custom linux users or bubblewrap and managing openbao is getting to be a lot to manage each harness by themselves!

New comment by tomrod in "Is 'Attention' the Essence of Intelligence?"

tomrod — Wed, 14 Oct 2026 05:02:59 +0000

This IS interesting and nifty!

New comment by tomrod in "AI is a tool and using it well takes skill"

tomrod — Tue, 13 Oct 2026 21:12:28 +0000

Oh... oh dear. As one of the few folks unspoiled by Silicon Valley inside baseball, that's somewhat horrifying.

New comment by tomrod in "A look at the math that OpenAI claims computes faster Fourier transforms"

tomrod — Tue, 13 Oct 2026 04:32:06 +0000

Does the apply to representation of complex numbers, which is not uncountably infinite and represented by floats?

New comment by tomrod in "Why aren't AI companies being charged for hacking?"

tomrod — Tue, 13 Oct 2026 04:06:44 +0000

I also listed several criminal options!

New comment by tomrod in "AI is a tool and using it well takes skill"

tomrod — Tue, 13 Oct 2026 00:17:40 +0000

I mean, I love rotary sanders too, but I'm not sure I'd call them the latent space of everything a human has ever written sitting in a harness of ever increasing complexity. That's not really my form of woodworking. But LLMs do fit on discs, even spinning HDDs on occasion (though I suspect it would be low tokens for second).

Mercury is incredible useful despite its hazards. Nitric acid is worse in negative affect despite its utility -- acid that rots your bones. When I visited fabs they were careful to point out which vats would burn me and which would make me instead lose the bone in an arm. Fortunately I survived the visits.

A rotary sander can damage plenty if you're not careful.

All these really are the same tune.

New comment by tomrod in "Why aren't AI companies being charged for hacking?"

tomrod — Mon, 12 Oct 2026 23:23:59 +0000

The FTC can fine for deceptive practices. There are a slew of vicarious criminal recklessness and liability approaches that charge an operator. Damage to infrastructure falls under public welfare and strict liability. There are state-level AI statutes (like Wyoming law that doesn't really allow "autonomous" defense) For civil damages, the CLAIM Act can be used to sue devs for negligince.

The law is much more broad than CFAA!

New comment by tomrod in "AI is a tool and using it well takes skill"

tomrod — Mon, 12 Oct 2026 23:16:36 +0000

HAH! Is that the same guy claiming misalignment to his personal politics makes a person "the AntiChrist" or some similar AI-induced psychotic slop?

New comment by tomrod in "AI is a tool and using it well takes skill"

tomrod — Mon, 12 Oct 2026 22:57:07 +0000

The first few times I used a rotary sander were similar to a slot machine too -- scary times for a bit there! Fortunately, and I think it applies to LLMs too, you can get more reliable results with experience.

New comment by tomrod in "AI is a tool and using it well takes skill"

tomrod — Mon, 12 Oct 2026 22:52:12 +0000

What do you think drives this bimodality?

New comment by tomrod in "Why aren't AI companies being charged for hacking?"

tomrod — Mon, 12 Oct 2026 21:37:20 +0000

Better question: why should CFAA be the entirety of law to be considered?

New comment by tomrod in "OpenFatigue™ – Open is the new closed"

tomrod — Mon, 12 Oct 2026 20:57:06 +0000

Have you put anything up on the internet using an AI prompt to build it? Would love to see it. Especially if it is clever or tongue in cheek.

New comment by tomrod in "'AI' Is Not a Tool by Rich Hickey"

tomrod — Mon, 12 Oct 2026 20:27:07 +0000

AI, as it currently exists in the marketplace, is used, engaged, and distributed to be a tool.

"AI" as it was known until ~2023 could also be useful as an algorithmic implementation to solve an objective.

Now that definition is squishy, but still relevant.

New comment by tomrod in "I Found a $90k Remote Code Execution Vulnerability in Google"

tomrod — Mon, 12 Oct 2026 18:42:31 +0000

Correct. The headline is incorrect with both the quoting of AI as well as the claim it isn't a tool.

You can hardly blame anyone for avoiding a clickthrough into an auth-walled, user-hostile site to read the point the headline made in an attempt to "gotcha"

New comment by tomrod in "Why aren't AI companies being charged for hacking?"

tomrod — Mon, 12 Oct 2026 18:35:57 +0000

That's not why the CFAA came about. It was just hard, not impossible.

But, let's go down the rabbit hole, eh? There are several legal theories under which an AI agent's operator can be charged:

- Direct perpetration, with the AI as the instrument. 18 USC section 2 (principals), and possibly wire fraud (18 USC Setion 1343). Under this, the law treats the defendent with using the AI as a tool (which the big AI firms argue anyway) which punishes the operator

- Criminal recklessness or even negligence - 18 USC Secion 1112 (Involuntary Manslaughter) - if someone died, the agent operator could be pursued this way

- Strict liability - 15 USC Section 1, states strict liability statutes. Certain regulatory offenses don't require intent. These specific statutes could apply to algorithmic price fixing or self-driving cars being reckless. The operator using the agents are liable for the interaction

- Probably the strongest is willful blindness (ostrich doctrine). Great common law precedence. This would be used to satisfy any intent/knowledge requirements like CFAA. You as an operator can't dismiss liability by just ignoring an AI is committing a crime.

- Vicarious liability - good commmon law precedence for corporate criminal liability. Agents acting in scope for an operator means the operator has liability - doesn't matter if that agent is human or AI-driven.

New comment by tomrod in "Why aren't AI companies being charged for hacking?"

tomrod — Mon, 12 Oct 2026 17:56:18 +0000

Yes, the legal code is actually larger than CFAA.

New comment by tomrod in "I Found a $90k Remote Code Execution Vulnerability in Google"

tomrod — Mon, 12 Oct 2026 16:19:36 +0000

Sure it is. And it's AI, not 'AI' -- not sure why the need to add quoting.

New comment by tomrod in "Why aren't AI companies being charged for hacking?"

tomrod — Mon, 12 Oct 2026 15:39:52 +0000

Eh. There is enough flexibility in the law for unintended damages to be prosecutable.