Hacker News: maleldil

New comment by maleldil in "Firefox is now the last major browser that still supports uBlock Origin"

maleldil — Sun, 16 Aug 2026 12:43:36 +0000

By "web features" you mean things Google decided they wanted and neglected to put through the actual process. This is why Firefox is important; the alternative is letting Google control the Web like Microsoft did.

New comment by maleldil in "AI has access to a vastly larger working memory than the human brain"

maleldil — Sat, 15 Aug 2026 21:51:58 +0000

Yeah, publishing in ML can be annoying because of that. What's wrong with "I thought this problem was interesting"?

New comment by maleldil in "Maximizing the value of your Claude Code sessions"

maleldil — Sat, 15 Aug 2026 21:21:57 +0000

Isn't 1+2 just reverse TDD?

New comment by maleldil in "Jason Arday, ex-Cambridge professor at centre of plagiarism row, found dead"

maleldil — Sat, 15 Aug 2026 21:00:23 +0000

Good luck getting accepted in another program.

New comment by maleldil in "Brazilian election filter in X For You timeline"

maleldil — Sat, 15 Aug 2026 19:42:57 +0000

Funny how someone implicitly reveals his political affiliation with a short sentence.

New comment by maleldil in "Text AI watermarks will always be trivial to remove"

maleldil — Sat, 15 Aug 2026 03:12:02 +0000

I agree completely. That's why I pushed back on "nothing to do" from the OP. Maybe it was a small contribution, but calling it "corruption" is ridiculous.

New comment by maleldil in "Hello, me. It's been a while"

maleldil — Sat, 15 Aug 2026 03:06:54 +0000

I never really understood how people can wear headphones all day. I have the XM5 headphones, and after a couple of hours, it starts hurting my head. What am I doing wrong? I also have the earbuds of the same generation, and they're much more comfortable, but the noise cancelling is less consistent.

New comment by maleldil in "DeepSeek Harness developer preview"

maleldil — Sat, 15 Aug 2026 02:56:21 +0000

Isn't distillation done through SFT by having the student model approximate the teacher's output distribution?

New comment by maleldil in "DeepSeek Harness developer preview"

maleldil — Sat, 15 Aug 2026 02:53:16 +0000

They were impersonating an AI. You wouldn't complain that a comic started impersonating a celebrity without saying "I'm going to impersonate X now" first.

New comment by maleldil in "DeepSeek Harness developer preview"

maleldil — Sat, 15 Aug 2026 02:49:21 +0000

Agree with sibling. If you're getting severely deteriorated results with structured output, you're probably doing something wrong. There's been some research on the impact of structured outputs on results distribution, and there are tradeoffs, but "do not work well at all" doesn't match the experience at large.

New comment by maleldil in "Text AI watermarks will always be trivial to remove"

maleldil — Fri, 14 Aug 2026 03:58:55 +0000

Verifying authorship is an ambiguous task. How much involvement is necessary to qualify for authorship? If someone reads the paper and gives some small feedback, is that enough? What if they were present in one meeting and raised a question that turned out not to be interesting? What if they have no clue about the work but helped with data validation?

I'd argue all of these could justify authorship, even if they're just in the middle of the author list. At least in NLP, which can be seen in the generally high number of authors in papers.

New comment by maleldil in "Text AI watermarks will always be trivial to remove"

maleldil — Fri, 14 Aug 2026 03:54:42 +0000

Peer reviews is largely done for free. In some cases, an author needs to commit to reviewing someone else's work for their work to be reviewed. However, this has led to an increase in LLM usage for review generation, even if conference/journal guidelines forbid it. I've seen nonsensical reviews from people who obviously haven't read the paper beyond an LLM-generated summary. Area chairs are supposed to catch this, but I imagine they're using LLMs too.

New comment by maleldil in "Mozilla says stricter antitrust measures against Google 'threaten Firefox'"

maleldil — Fri, 14 Aug 2026 03:39:51 +0000

And when the Google money runs out, what will they do?

New comment by maleldil in "Mozilla says stricter antitrust measures against Google 'threaten Firefox'"

maleldil — Fri, 14 Aug 2026 03:37:57 +0000

Maintaining a Web browser is hard. I'm not convinced that an open source community would be able to continue the development.

New comment by maleldil in "Accelerating GPT-5.6 Sol Ultrafast"

maleldil — Fri, 14 Aug 2026 02:08:37 +0000

Reasoning models will frequently backtrack and re-assess what they've said so far. That's one reason test-time scaling is so powerful.

New comment by maleldil in "NP-Overrated"

maleldil — Thu, 13 Aug 2026 23:55:47 +0000

OR has some very interesting algorithms for very interesting problems. I don't do research on it anymore, but going to OR conferences was always interesting. Good mix of practitioners, researchers and end users. I'm way behind SOTA now, but I have a soft spot for evolutionary algorithms.

New comment by maleldil in "NP-overrated"

maleldil — Thu, 13 Aug 2026 23:51:40 +0000

How often have you encountered code that adds a timeout to regex matching?

New comment by maleldil in "NP-Overrated"

maleldil — Thu, 13 Aug 2026 23:50:21 +0000

I think that Swift is the exception that proves the rule. The fact that it is notorious for giving up on certain type checks indicates that it isn't a problem for most languages.

New comment by maleldil in "DeepSeek V4 Pro 0813"

maleldil — Thu, 13 Aug 2026 04:08:15 +0000

Your tool is giving "Error: Gist API returned 403"

New comment by maleldil in "Why Target Common Lisp for Code Generation?"

maleldil — Thu, 13 Aug 2026 03:58:52 +0000

REPLs aren't debuggers.