Debian has begun voting on the future of AI/LLM contributions | Hacker News Reader

Debian has begun voting on the future of AI/LLM contributions

lists.debian.org

> After the voting period there will be a record of all the votes without the name of the voter. It will instead contain a cryptographic hash. You will receive a secret after you have voted that can be used to calculate that hash. This allows you to verify that your vote is in the list. This secret is sent in an encrypted mail.

I've always wanted this for mail-in voting. E.g. print a random secret on the voting form I receive per mail. Afterwards publish the complete vote-count on the web, with the choice and a hash of the secret. So I can check that my vote was recorded, and only once, and correctly.

I understand there might be a lot of reasons for why it could fail in practice, but it would be a start.

I think the main problem with this is that you can be paid/coerced to vote for someone and be able to easily prove that you voted for the person they wanted. It’s easier to vote for the person you actually wanted to vote for, and maintain plausible deniability, if you go in person to a voting booth.

Could this be mitigated by allowing voters to vote multiple times and receive multiple confirmations, but only count the last vote for each person? That way an individual could prove that they voted a certain way to a third party, but there would be no way for the third party to know whether it was their final vote. Meanwhile the individual knows that they spoiled their previous vote or votes and that their final vote is included.

You’d have the same issue as you’d need to be able to verify that your last vote was being counted, for displaying your vote back to you to have any value.

A scheme to be able to show coercive parties fake votes would be known to them immediately.

This wouldn't solve anything. The person doing the coercion would be able to store the hash and double check after the election if the vote was still the way they wanted.

the question is, what's the worst situation for society.

People not trusting the votes being counted properly or people being coerced paid to vote some way and this making it more valuable.

A plausible solution would be to significantly delay the distribution of the record. i.e. if every election was only printed out 2-4 years after the fact, it could give people confidence that the votes are being counted properly (as otherwise they would be found out eventually), but significantly reduce the incentive for the bad behavior (if one things this actually significantly incentivizes it) because it will be a long time before they can validate if what they were given is true.

I don't take this concern seriously.

You can be paid/coerced to vote for someone even if there's total secrecy. For example, you can take your cellphone in and take a picture of your filled out ballot to collect a payment.

Unless you want to start the confiscation of cellphones before voting in person, the risk of a coerced or bought vote will always exist.

In Argentina at least, you can get a fine[1] for taking a picture of your ballot. Certainly a hard law to enforce though.

[1]https://www.perfil.com/noticias/politica/elecciones-2025-de-...

> For example, you can take your cellphone in and take a picture of your filled out ballot to collect a payment.

And then you can go up to the election workers and tell them you have a spoiled ballot and want a new one, which they have to give you. That's one reason such schemes don't work right now.

Source: I volunteer as an election worker every year in Iowa.

In Texas, the election judges are supposed to stop you if you pull out a cellphone or any device. Paper is all you are allowed to use during the voting process.

In-person, high-risk rule-breaking that's hard to validate has a tendency to scale differently from say, pasting a value into a lottery/payment app.

I mean, compare the number of people who've flown over Barbara Streisand's property to take a picture of her house to the number of people who had a copy of said picture in their browser cache.

You can read more about this peculiar scaling factor of the internet in Bill Gates 1995 book The Road Ahead.

The unfortunate reason this isn't doable in the US anyway, and any time someone has a great idea on how to improve elections this inevitably comes up: preventing coersion. It's difficult to balance transparent voting with ensuring that you can't prove which way you voted in exchange for money or favors.

You can take a picture of your filled out ballot with a cellphone. I don't see how there isn't already a very wide open door for coercion/vote purchasing. It's not 100% fool proof, but you don't need that to sway an election, you just need enough people to follow through.

But even with that, it seems like the better course of action is criminalizing such coercion (which is already illegal). Anyone doing something like that on a scale large enough to matter would be caught rather quickly.

There’s places where taking photos inside the voting box is illegal for this exact reason.

This could be prevented if the government offered photos of ballots with votes for download. But generative AI can also fake it well enough these days. (Of course, this is less of a concern in countries where taking photos of ballots is not allowed.)

For me, the main reason that I'm against all high-tech voting solutions is that a less-than-gifted citizen must be able to check that, at least for their polling station:

- Their vote, and everyone else's, was counted exactly once.

- The number of ballots that are counted is the same as the number of people who voted.

- Votes are anonymous.

All that is easily done with a clear ballot box, a signing registry, and public counting. Adding technology to that just makes the process more difficult to trust IMO.

As for myself, you're going to have to convice me that the software + hardware that's running on election day is exactly conformant to its published source code (the source code is available, right?)

For something like voting on decisions for some open source project, sure.

For actually voting on representatives it would be an absolute disaster.

Don't blame me, I voted for Jia Tan!

It's not secret to the administrator though, as they could know who was assigned what secret. I think it could work like: before voting, you go to an office where they verify your identity and let you randomly pick envelope containing unique number from a box. After voting all votes are published and you can verify your vote was counted, but nobody else knows your number.

The choices are oddly and not consistently worded. "Ban LLM contributions from Debian via Social Contract", "A cautious approach to generative AI", "Debian is created by humans" and "Avoid the use of LLM: climate destruction is a deal breaker" all in one list? Sounds like several people wrote them and did not agree on the form.

This is the way general resolutions in Debian generally work. There's a lot of pre-vote discussion where different people and groups propose and draft their preferred options, to produce what they feel is the best expression of their position. If you can get enough seconders for your proposal it goes on the ballot, with the text you chose and your seconders agreed with. There isn't any requirement for the options to use consistent terminology or avoid overlap, and no overarching editorial process to impose consistency (which would be tricky anyway in cases like this where terminology choices and overall framing can be part of the disagreement between people.)

Thank you for a valuable response; I did not know this.

> Sounds like several people wrote them and did not agree on the form

Of course several people wrote different position statements to vote on.

Did you expect one Debian developer to cook up half a dozen possible stances he is not even personally invested in? Are you constantly surprised that election programs of different parties are worded differently (and use different fonts)?

Obviously not, you misunderstood my point. And the election programs comparison makes no sense.

Dude, chill with the strawman and the aggression. Bad day?

Fwiw, I perceive no aggression in their post.

There is no aggression, and "dudeing" me makes me disregard you immediately anyway.

If we’re disregarding people based on how they express themselves, Keybase is owned by Zoom now…

Tried to decipher this but seems a bit convoluted. If the Kernel now allows LLM code, then what will Debian do if they ban LLM code? Choice 7 and choice 8 seem pretty clear that the project simply can't continue under those circumstances.

Neither choice 7 nor 8 attempt to forbid packaging of any upstream project with LLM contributions like the kernel.

Choice 7 says "The proposal does not cover indirect contributions, i.e. those originating from upstream works.". It's only making requirements on Debian work specifically.

Choice 8 says "We also acknowledge that LLM usage can be hard if not impossible to detect and that Debian as a distribution cannot really impose LLM policies on other projects we package and distribute. Therefore this text is just a position statement." It doesn't impose any requirements at all on anybody, even for debian specific work: it just says "we'd prefer it if you don't use this technology".

Well, they could try to prove the correctness of that LLM-based kernel code.

If it passes, then they could use it, no?

Proposal A would of course be the best option, but requires a two thirds majority for... some reason. Proposals B, C, D, E are all defeatist policies that appeal to a "here to stay" narrative. Proposal G is potentially workable but has so many carve-outs to be practically useless, leaving only Proposal H as a workable policy if A fails to pass.

How conservative.

Is that a bad thing? Not everyone needs to chase VC money

TIL being climate conscious is "conservatism"....

Reply on news.ycombinator.com