martin-t

83 ms·

691.
by martin-t 2y ago
The issue with this approach to moderation is that it targets posts based on visibility of "undesired" behavior instead of severity. For example, many manipulative tactics (e.g. the fake sorry here, responding to something else th
692.
by martin-t 2y ago
[flagged]
693.
by martin-t 2y ago
Sounds like the tests would benefit from comments explaining what they are testing and why. Just a few days ago there was an article from an experienced dev that after multiple decades he thinks you simply can't put too many comments i
694.
by martin-t 2y ago
This article has been submitted at least twice already and got flagged into being dead immediately, despite all comments agreeing with it and despite users vouching for it. There is clearly a large group against free discussion of this topi
695.
by martin-t 2y ago
And now this one is dead too. In case I mange to post this before it's dead again, this article is from a military historian and has a lot more substance: https://acoup.blog/2024/10/25/new-acquisitions-19
696.
by martin-t 2y ago
This article was submitted earlier [0] but flagged and dead without a single comment disagreeing with its contents. I vouched for it and it was dead again before I managed to comment. - Anonymous voting does not work. People who have a beli
697.
by martin-t 2y ago
TBH I'd say if you're looking for a pure Rust engine, it's production ready already. I used it a year or two ago and it was reliable, fast and had most of what I needed. I asked for some fringe features like improvements to t
698.
by martin-t 2y ago
Yesterday I had a very similar argument. Maybe I haven't been paying attention until now but this is the first time I see somebody else summarize it so clearly. I've come to the conclusion that value (you call it wealth) is create
699.
by martin-t 2y ago
I get replies like this all the time and I am starting to conclude that most of the people arguing that LLMs are intelligent/reasoning are either completely clueless about what those words mean or are arguing in bad faith.
700.
by martin-t 2y ago
The nuance here being that this only proves additional censorship is applied on top of the output. It does not disprove that (sometimes ineffective) censorship is part of the LLM or that censorship was not attempted during training.
701.
by martin-t 2y ago
So, first, this is factually wrong, there are obviously new ideas, often rediscovered independently. Second, seeing things in black and white is a fallacy. The amount obviously matters and what also matters is the _human_ work on top of the
702.
by martin-t 2y ago
It's about proportions. 1) The current system is that the rich get richer faster than the poor. 2) You're proposing a system where everybody gets richer at the same rate at the best of times (but probably devolves into case 1 IMO)
703.
by martin-t 2y ago
Both your arguments here seem to boil down to preferring collective good over individual good even at the expense of individuals and even against the individuals' wishes. That's a totalitarian approach and I cannot and will not su
704.
by martin-t 2y ago
What is "free code"? Most code is either proprietary (though sometimes public), under a permissive license or under a copyleft license. The only free code is code which is in the public domain. You could make separate versions of
705.
by martin-t 2y ago
> Patenting algorithms is like patenting thoughts. OK, I agree there, I should have written "function" or "module" something similar. Something that takes nontrivial amounts of work and although it is based on some ge
706.
by martin-t 2y ago
I knew about that case (though I don't know the specifics of how sophisticated the LaMDA model was at the time, I don't know if it ever was available online so I could try it). AFAICT Blake Lemoine was not concerned with copyright
707.
by martin-t 2y ago
There are 2 separate metrics - obviousness (=provability) and reality. Courts operate on provability (and for good reasons). However, reality is I have used someone else's work and pretended it's my own. Now, ironically, there are
708.
by martin-t 2y ago
> I might consider it if any of those claims were true. Please step back and consider what you are replying to. > LLMs are used to launder my code If an LLM was trained only on AGPL code, would it have to be licensed under AGPL? Would
709.
by martin-t 2y ago
> AFAIK under EU law only blatant copies are an infringement Laws generally don't encode what is right but a compromise between the state's interests, lobbyists and the general population making enough ruckus if too unsatisfied
710.
by martin-t 2y ago
I wouldn't say that. My point is that any situation where person A puts in a certain amount of work (normalized by skill, competence, etc.), person B uses person A's work, puts in some work of his own but less than A, then gets mo
711.
by martin-t 2y ago
What I describe isn't the current system, it's the reform I propose. The current system sadly often is you can't self-publihsh because "lol no marketing" so you sell your rights for a one-time fixed payment and the
712.
by martin-t 2y ago
It's author's life plus 70 years, if you meant that. And TBH I am mostly interested in the author's life part anyway. If it was possible to train AI models on just the public domain, them I am sure ML companies would have bec
713.
by martin-t 2y ago
You have a point but 1) I don't think being able to query them and reconstruct input 1:1 are requirements. If i build a shitty db with a buggy query language that retrieves incomplete data and occasionally mixes in data i didn't a
714.
by martin-t 2y ago
I am writing a master's thesis and notice somebody has written one containing a chapter than I also need to write. I copy paste the chapter but replace every word with a synonym. Did I break copyright? Did I commit plagiarism?
715.
by martin-t 2y ago
None of the people who claim LLMs are intelligent and "persons" argue for giving them legal personhood and human rights. Telling.
716.
by martin-t 2y ago
> And if one wants to train an LLM, let them: at its essence it's just a person If I gzip a couple hundred thousand books and distribute them freely, can I also claim it's just a person who has read those books and avoid a mass
717.
by martin-t 2y ago
I haven't noticed a particular date but yes, my posts describing a fair system of compensation get downvoted incredibly fast. At least I sometimes also get replies but many of them use fallacious arguments to the point of feeling like
718.
by martin-t 2y ago
Not ridiculous at all. If LLMs (can we please stop calling it AI?) can produce correct factual statements (for example about historical events), then the data is clearly present in the model in some (compressed) form. The only question then
719.
by martin-t 2y ago
> pay artists to create something AI free / copyleft free I still don't think that's enough to be fair. If their work is used to produce value ad infinitum, them any one-time payment is obviously less than what they deserv
720.
by martin-t 2y ago
1) For starters, ML companies clearly go through enormous amounts of money, the rich people in control of them very obviously get compensated quite generously, if you're into euphemisms, which the people who built the "training da

More ›