Hacker News: kanzure

New comment by kanzure in "Codex is now in the ChatGPT mobile app"

kanzure — Fri, 15 May 2026 14:39:56 +0000

I haven't found that to be an issue. Just say what revisions you want. Once you're done, paste it into an LLM to clean it up into a usable prompt.

New comment by kanzure in "Codex is now in the ChatGPT mobile app"

kanzure — Fri, 15 May 2026 13:52:33 +0000

It's a lot easier to write long messages on a phone with something like Whisper.

New comment by kanzure in "GitHub appears to be struggling with measly three nines availability"

kanzure — Mon, 23 Mar 2026 14:11:44 +0000

I've always been worried about their backend changing and somehow named tags with a previous commit hash working for an attacker to give something you didn't expect for the commit hash.

New comment by kanzure in "Give Django your time and money, not your tokens"

kanzure — Tue, 17 Mar 2026 15:14:59 +0000

I like the idea of donating money instead of tokens. I think django contributors are likely to know how to spend those tokens better than I might, as I am not a django core contributor.

Some projects ( https://news.ycombinator.com/item?id=46730504 ) are setting a norm to disclose AI usage. Another project simply decided to pause contributions from external parties ( https://news.ycombinator.com/item?id=46642012 ). Instead of accepting driveby pull requests, contributors have to show a proof of work by working with one of the other collaborators.

Another project has started to decline to let users directly open issues ( https://news.ycombinator.com/item?id=46460319 ).

There's definitely an aspect here where the commons or good will effort of collaborators is being infringed upon by external parties who are unintentionally attacking their time and attention with low quality submissions that are now cheaper than ever to generate. It may be necessary to move to a more private community model of collaboration ( https://gnusha.org/pi/bitcoindev/CABaSBax-meEsC2013zKYJnC3ph... ).

edit: Also I applaud the debian project for their recent decision to defer and think harder about the nature of this problem. https://news.ycombinator.com/item?id=47324087

New comment by kanzure in "Don't make me talk to your chatbot"

kanzure — Wed, 04 Mar 2026 00:26:10 +0000

It would be nice if more businesses embraced email instead of requiring phone calls for basic tasks. Imagine how much more productive we could be if we could just send off a quick email with the information and questions.

Instead, what we're likely going to get are "voice agents" calling each other when we could have just used email instead...

New comment by kanzure in "Show HN: Open-Source Article 12 Logging Infrastructure for the EU AI Act"

kanzure — Tue, 03 Mar 2026 21:30:38 +0000

Anyone can generate an alternative chain of sha256 hashes. perhaps you should consider timestamping, e.g. https://opentimestamps.org/ As for what the regulation says, I haven't looked but perhaps it doesn't require the system to be actually tamper-proof.

New comment by kanzure in "New accounts on HN more likely to use em-dashes"

kanzure — Thu, 26 Feb 2026 16:19:12 +0000

Having multiple different distribution channels can solve that problem. Advertisers cannot monopolize all distribution channels simultaneously because of the costs involved (it would be like someone trying to buy the whole economy).

New comment by kanzure in "New accounts on HN more likely to use em-dashes"

kanzure — Wed, 25 Feb 2026 21:32:35 +0000

Another option instead of using identity is to use proof of work or hashcash such that anyone who thinks a comment is valuable can use some hash rate to upvote it. It doesn't matter how the content was generated, only that someone thought it was important, and you can independently verify this by checking how much hash effort went into hashing for that comment. This also does not require any identity either.

New comment by kanzure in "AI Usage Policy"

kanzure — Fri, 23 Jan 2026 11:19:57 +0000

Another project simply paused external contributions entirely: https://news.ycombinator.com/item?id=46642012

Another idea is to simply promote the donation of AI credits instead of output tokens. It would be better to donate credits, not outputs, because people already working on the project would be better at prompting and steering AI outputs.

New comment by kanzure in "Tldraw pauses external contributions due to AI slop"

kanzure — Fri, 16 Jan 2026 01:43:56 +0000

That's interesting; another project stopped letting users directly open issues: https://news.ycombinator.com/item?id=46460319

New comment by kanzure in "Show HN: OSS AI agent that indexes and searches the Epstein files"

kanzure — Wed, 14 Jan 2026 15:35:29 +0000

Wait, is abliteration actually just a branding trick? That doesn't sound correct.

New comment by kanzure in "Why users cannot create Issues directly"

kanzure — Fri, 02 Jan 2026 03:10:22 +0000

I proposed something similar for bitcoin: https://gnusha.org/pi/bitcoindev/CABaSBax-meEsC2013zKYJnC3ph...

New comment by kanzure in "Generative Optogenetics"

kanzure — Tue, 16 Dec 2025 00:05:11 +0000

Oh it's much older than 8 years... here is one such proposal from 2011: https://groups.google.com/g/enzymaticsynthesis/c/6GZT8zFNOfo...

I myself was bitten by a radioactive grad student in 2008 that was obsessed with this idea at the time, and have since learned that almost every major household name lab PI has thought about this in one form or another.

New comment by kanzure in "Super Mario 64 for the PS1"

kanzure — Wed, 10 Dec 2025 21:05:10 +0000

To me this sounds like a computer-generated voice for obvious pro-privacy reasons for this kind of project. If it bothers you, then maybe work on better voice synthesis tech! I assume it sounds not-leading-generation because it was locally rendered but I could be wrong.

New comment by kanzure in "Developers are choosing older AI models"

kanzure — Wed, 05 Nov 2025 14:36:57 +0000

You can also switch between models with aider https://aider.chat/

New comment by kanzure in "Public trust demands open-source voting systems"

kanzure — Tue, 21 Oct 2025 18:01:06 +0000

Why stop at software? Open-source software is a good idea in election systems. The principle could be better generalized as an "open" (copyleft licensed) process for the entire system, regardless of whether the election system is implemented as software or not.

Anyone who talks about election security should be required to spend at least a few moments walking around Defcon in the election machine hacking village. Even absent electronic voting machines we still need to apply that same level of rigor to security across all domains of the election system no matter what format is used.

More fundamentally, the epistemic meaning of a ballot, a vote, or an option on the ballot, how options are even decided for inclusion or their exclusion, which outcome deciding algorithms are used, and how "the result" is interpreted by society or implemented by a political agent is deeply confused. The vote itself has very little resemblance to what actually happens. Such things likely cannot be formally specified anyway. Massive amounts of ambiguity, noise, error rate, and insecurity are to be expected in these kinds of systems. So what then are we even doing with all this? I am not referring to what we say we are achieving, or what we say we are intending to achieve, but rather what kind of actual outcomes be can supported by careful engineering of all these components?

Blockchain is no solution here. See:

"Going from bad to worse: from Internet voting to blockchain voting" https://www.dci.mit.edu/s/VotingPaper-RivestNarulaSunoo-3.pd...

New comment by kanzure in "Why are there so many rationalist cults?"

kanzure — Tue, 12 Aug 2025 21:44:27 +0000

Here are some other anti-lesswrong materials to consider:

https://aiascendant.com/p/extropias-children-chapter-1-the-w...

https://davidgerard.co.uk/blockchain/2023/02/06/ineffective-...

https://www.bloomberg.com/news/features/2023-03-07/effective...

https://www.vox.com/future-perfect/23458282/effective-altrui...

https://qchu.substack.com/p/eliezer

https://x.com/kanzure/status/1726251316513841539

New comment by kanzure in "Show HN: Autumn – Open-source infra over Stripe"

kanzure — Wed, 25 Jun 2025 15:21:32 +0000

Huh, weird. I don't think I have ever wanted to login to Stripe and take manual actions. Usually I hook up Stripe events to actions in my applications, like "The user subscribed, wire them up to the drip lifecycle" or "The user unsubscribed, remove them from a certain marketing list and try to schedule an exit interview" or other hook-ups.

New comment by kanzure in "Show HN: Autumn – Open-source infra over Stripe"

kanzure — Tue, 24 Jun 2025 17:11:38 +0000

Hi, your docs say that users don't need state syncing, but when using Stripe you do need state syncing or to ingest the Stripe events. I also don't see any information in the docs about handling e.g. chargebacks or other events and listening for (or otherwise syncing against the history of) those events. I'm a little confused - why would I want to not have that? I could be misunderstanding this project though!

New comment by kanzure in "Show HN: A web browser agent in your Chrome side panel"

kanzure — Sun, 18 May 2025 20:21:18 +0000

possibly something like https://github.com/romansky/dom-to-semantic-markdown could also help for this use case.