Hacker News: cobbal

New comment by cobbal in "Show HN: Let your AI agents paint big arrows, boxes and text on your screen"

cobbal — Fri, 09 Oct 2026 16:22:19 +0000

Finally we've built the reverse centaur from Cory Doctorow's classic novel "Don't Become the Reverse Centaur"

New comment by cobbal in "RuneScape's Position on Gen AI"

cobbal — Sun, 04 Oct 2026 17:28:53 +0000

Why is it unrealistic? We have decades that prove people can make games without AI. Is the unrealistic part just expecting people to be honest about whether they used it or not? We somehow manage to have the food industry mostly honest about what they put in.

New comment by cobbal in "Tokens too cheap to meter"

cobbal — Wed, 23 Sep 2026 17:39:35 +0000

I am now imagining a future where developers buy fancy "grep cards" for their machines. I don't hate it.

New comment by cobbal in "XCancel service is suspended until further notice"

cobbal — Mon, 14 Sep 2026 17:24:50 +0000

Twitter promised an open publishing platform with extensive API access. They're the ones who broke the (social) contract, not us.

New comment by cobbal in "Tao: Open math problems being non-renewably mined by AI"

cobbal — Wed, 09 Sep 2026 04:07:42 +0000

This is definitely true in an information theory sense: having more knowledge is always better than less knowledge. However, it may not be true in math as a social human endeavor, and having answers without interesting paths to get there may not expand human mathematics in the same way.

If Fermat had a book with larger margins, would Weil have devoted so much time to proving the Taniyama-Shimura conjecture? No one can say.

New comment by cobbal in "Stopping the smart TV from being used against you"

cobbal — Fri, 28 Aug 2026 22:11:17 +0000

It does sound totally ridiculous. Sadly it may not be. Previously discussed: https://news.ycombinator.com/item?id=48956688 "LG monitors silently install software through Windows Update without consent".

You can argue if it actually installs the software or just nags the user to install it, but neither are things microsoft would allow a monitor "driver" to do if they had any integrity.

New comment by cobbal in "Software engineering is about managing complexity"

cobbal — Thu, 27 Aug 2026 15:21:25 +0000

Lost me at the first assumption. People can argue about how useful AI is, but it's obviously not essential because we somehow managed to write code without it a few years ago. I would even say the code was better back then.

The two tasks of writing code and engineering software cannot be separated without damaging the integrity of the mental model of the engineer. Having architects who didn't interact with the code always produced map/territory mismatches.

New comment by cobbal in "Claude: Elevated errors across all models – Resolved"

cobbal — Wed, 29 Jul 2026 20:21:20 +0000

AI is a technology for removing nines from uptime, and adding them to Sturgeon's law.

New comment by cobbal in "Now is the time to give LLMs access to the ACM digital library"

cobbal — Tue, 28 Jul 2026 17:31:12 +0000

ACM agrees I think: https://dl.acm.org/openaccess

> Beginning January 2026, all ACM publications and related artifacts in the ACM Digital Library will be made open access.

New comment by cobbal in "Claude Is Not a Compiler"

cobbal — Tue, 21 Jul 2026 16:17:41 +0000

Determinism isn't the main hurdle. It would be a moderate effort to engineer a deterministic claude that would be no more useful than the current one.

The key characteristic of a compiler is that it takes a source language with a well-defined meaning to a target language with a well-defined meaning.

The reason claude fails to be a compiler is that its input language (english) can't be precisely defined. (And even if that were possible, it would very often fail to preserve the meaning when "compiling")

New comment by cobbal in "FCC approves test of space mirror to light night sky"

cobbal — Sat, 11 Jul 2026 16:05:41 +0000

You may have been thinking of this one instead https://what-if.xkcd.com/141/ . Maybe we should put Randall on the supervillain watchlist.

New comment by cobbal in "Leaking YouTube creators' private videos"

cobbal — Sat, 04 Jul 2026 21:13:57 +0000

This is a case of lethal trifecta. This particular one can be fixed by either not giving the AI private data, or by removing the exfiltration opportunity. Why does the comment-summary bot need access to your private video ids? Why does it need to be able to output links?

Most cases of prompt injection are harder to fix, and the success of the products they occur in relies on engineers who should know better sticking their heads in the sand about security risks.

New comment by cobbal in "Statement on US government directive to suspend access to Fable 5 and Mythos 5"

cobbal — Sat, 13 Jun 2026 01:01:12 +0000

Sadly, I suspect this will be the best piece of marketing they could ever hope for. "It's so advanced the government made us add extra security* to stop hackers!"

*(ask it in a more stern voice)

New comment by cobbal in "Show HN: Continue? Y/N: A 60-second game about AI agent permission fatigue"

cobbal — Thu, 28 May 2026 17:53:04 +0000

Here's the threat model I (a luddite) use to evaluate these. The claude code harness can be mostly trusted, the model cannot be trusted because it is exposed to untrusted data from the internet, and there is no separation of data/code in an llm [0][1].

I want to avoid running untrusted code on my local machine, because it could steal secrets, install malware, etc.

Since the model is allowed to write without restriction (I think) to the project directory, anything in the project directory is also untrusted. Running standard commands from the system is fine, as long as you know what those commands are going to do. Running anything from the local directory should be avoided because the code is untrusted.

This is just one security model, there are many others! If a person is running claude in a stronger sandbox, that changes the model considerably. What threat model do you use to evaluate whether an agent's actions are safe?

[0]: https://www.schneier.com/essays/archives/2024/05/llms-data-c... [1]: https://simonwillison.net/2025/Jun/16/the-lethal-trifecta/

New comment by cobbal in "Show HN: Continue? Y/N: A 60-second game about AI agent permission fatigue"

cobbal — Thu, 28 May 2026 15:49:13 +0000

That's funny. It told me that blocking "npm run build" was the wrong answer. Maybe it doesn't really under The threat model.

New comment by cobbal in "Stop Advertising in Your Commits"

cobbal — Tue, 26 May 2026 19:54:44 +0000

If co-authorship implies holding rights, then what gives the "primary author" who just prompted for the code the right to add their own name?

New comment by cobbal in "AI has a multiplying effect on existing technical skills"

cobbal — Fri, 22 May 2026 17:25:54 +0000

It makes up for that by also multiplying the productivity and happiness of the people who have to review your code by 0.5 too.

New comment by cobbal in "What Do Gödel's Incompleteness Theorems Mean?"

cobbal — Thu, 21 May 2026 16:34:29 +0000

I think that's selling the theorems a little short. A math system with arithmetic is equal to, or more powerful than, a computer. For an example, even classical logic comes with the law of excluded middle that can say (internally) if a program halts or not. Incompleteness applies to all the stronger systems as well.

New comment by cobbal in "Union types in C# 15"

cobbal — Wed, 08 Apr 2026 18:24:19 +0000

Null doesn't. `union(Int?, String?)` will only have 1 type of null, unlike a proper discriminated union.

New comment by cobbal in "Show HN: Stop paying for Dropbox/Google Drive, use your own S3 bucket instead"

cobbal — Tue, 07 Apr 2026 15:50:36 +0000

Take your pick :) https://www.templatemaker.nl/en/