martin-t

53 ms·

721.
by martin-t 2y ago
So you think all work produced with the help of LLMs should by required to be open sourced under a copyleft-like license? That is actually an intriguing idea and at least aligned with the reasons I use AGPL for my code.
722.
by martin-t 2y ago
I am surprised by how quickly I get downvoted to hell any time I try to propose that people should be compensated for their work fairly. And I get replies like the above, full of fallacies, which basically feel like trolling. So first, than
723.
by martin-t 2y ago
The issue is treating "AI" as something special. It is just derivative work. They are called large language _models_ for a reason. They are just statistical models of existing work. If anybody seriously thought they were intellige
724.
by martin-t 2y ago
Exactly. It's impractical to distribute compensation (and credit) fairly so it's very easy for those who profit to say "we can't", their their hands up and keep profiting. Doesn't make it right. It's just
725.
by martin-t 2y ago
I have trouble believing this argument is in good faith. 1) (minor nitpick) I don't see how HN is in the same category as GitHub or Stack Overflow. 2) "sometimes by license" or "free by license" imply you don't
726.
by martin-t 2y ago
Even if they are small enough to be run be individuals, this still doesn't solve the issue of profiting from someone's work for free. Many of us publish our open source work under GPL or AGPL with the intention that you can profit
727.
by martin-t 2y ago
I was hoping the article would propose the opposite: if you train LLMs on copyrighted data, you owe the author a part of your income from it. How big should be determined by courts but probably proportional to the amount of data. There&#x27
728.
by martin-t 2y ago
Humans run on hardware that is both faulty and limited in terms of speed and memory. They have a better "algorithm" how to use the hardware to compensate for it. LLMs run on almost perfect hardware, able to store and retrieve enor
729.
by martin-t 2y ago
Comparisons to humans are ultimately misleading because 1) humans are not general intelligences most of the time, 2) humans run on incredibly faulty hardware. 1) Attention is limited. Human reasoning is slow. Motivation is limited. System 1
730.
by martin-t 2y ago
I am not gonna pretend to have one true definition but there are hopefully components of reasoning we can agree on: - The ability to hold a collection of beliefs/statements and perform operations on them such as noticing one contradict
731.
by martin-t 2y ago
At first I was curious if OP tried arguing with it or telling it the correct answer. If he just asked the question without any further interaction, it would still have no data that it generated the wrong answer. But as another comment says,
732.
by martin-t 2y ago
Thank you, this is a perfect argument why LLMs are not AI but just statistical models. The original is so overrepresented in the training data that even though they notice this riddle is different, they regress to the statistically more lik
733.
by martin-t 2y ago
> In very localised conditions, largely due to the state's neglect. I didn't mean just materialized physical violence against a person's body. It is still violence when it's implied/threatened, when it's tar
734.
by martin-t 2y ago
That's the second difference i mention. Organized crime is able to wield more violence than normal individuals so it has more power over them. I perhaps mistakenly used the word "certain" to describe state violence. I tried t
735.
by martin-t 2y ago
I believe most of what you're referring to are different situations such as people acting on impulses - either not considering the outcome or being resigned to it. Another key difference is usually those take place between individuals
736.
by martin-t 2y ago
I would say pacified instead of hopeless. All power in the real world comes from violence. This is increasingly a taboo to say but when you think about it, it becomes obvious. How do you put someone in prison without the ability to physical
737.
by martin-t 2y ago
Most of what generative models produce is shit so they have to produce a lot in hopes _some_ of it is OK-ish. It's also about responsiveness. LLMs produce junior-level quality of code at a rate of hundreds of lines per minute. I need i
738.
by martin-t 2y ago
Not just that but in the medieval world, you needed people to control people. With AGI, you can detect signs of revolt programmatically across the entire population and, well, take care of the "problems". And I have not seen anybo
739.
by martin-t 2y ago
> Pretty much everyone has been trained by society to follow the rules above all else as a strong moral imperative, no matter how stupid Oh, you are "aware", for lack of a better word. It's so depressingly rare to even see
740.
by martin-t 2y ago
I prefer a third approach - chaining method calls via `.method()`. C#'s LINQ extension methods and Rust's iterator methods do it and I find it to be the best of both worlds most of the time.
741.
by martin-t 2y ago
Fair point, i understood the quote differently (as promoting respect for the law even if unjust)
742.
by martin-t 2y ago
It's a tug of war between institutional capacity (people's inability to act) and injustice (people's willingness to act even if they will face negative consequences).
743.
by martin-t 2y ago
Yep, if you boil the frog too fast, it will notice.
744.
by martin-t 2y ago
I cannot agree with the quote for 2 reasons: 1) That person has respect for justice, not law. These are separate concepts and people need to see them as separate. 2) It reinforces the view that by making yourself a victim (of the state), yo
745.
by martin-t 2y ago
Not just vote him out, that solves all his future murders. Depending on when he was votes out, he still needs to be punished for all his past murders.
746.
by martin-t 2y ago
Yes but try to apply the same rules to real life and people will laugh at you. How many times do you see the hero having to choose between killing the villain and leaving him to the legal system (not justice system) to give him a lesser pun
747.
by martin-t 2y ago
The difference is in the past most discourse was by mouth - impossible to censor and eavesdropping was costly because you needed a person to do it. These days more and more discourse is online - heavily censored by platform rules which are
748.
by martin-t 2y ago
But that _have_ learned. They are systematically making violence a taboo the way sex was in the past. If all people think violence is wrong and immoral instead of just against the rules they have written ("laws"), then they can ke
749.
by martin-t 2y ago
Why would it work though? If somebody wants to actually detect and censor talking about rape, they certainly have the technological capability to also censor all the intentional misspellings. IMO, the only case when this works is when an or
750.
by martin-t 2y ago
There's a misconception that systematic harmful behavior has to be organized. The truth is that many antisocial people generate a strong cult following and these people protect the abuser. Given Altman's followers are likely to be

More ›