back
user profile

artninja1988

1,865karma·554submissions·August 15, 2023
recent activity (554 total)
comment
Good man
2y ago·view thread
comment
If datasets are not shared externally, like how deepmind etc. does it, it is much more of a grey area. Training may well end up being fair use. So yes, the efforts of Shawn compiling books3 just level…
2y ago·view thread
comment
I mean the pile, which books3 is only a subset of, is only about 850GB big. Raw text doesn't take up much space at all
2y ago·view thread
comment
Hope someone hosts it can't be taken down by imaginary "property" DMCA requests
2y ago·view thread
comment
Would also like to know this. Can't seem to find an open source tts engine that works on mobile to read muh books
2y ago·view thread
comment
Well, it did make you and me klick...
2y ago·view thread
comment
Why not
2y ago·view thread
comment
4) The WGA reserves the right to assert that exploitation of writers’ material to train AI is prohibited by MBA or other law What does that mean
2y ago·view thread
comment
Sounds a lot like discriminative hiring. Isn't that illegal in the us?
2y ago·view thread
comment
I don't think all the garbage in their training data would be worth that much :)
2y ago·view thread
comment
Man I the part about having to get a license to release a foundational model really friggin sucks and I hope it doesn't make it to the final version of the text
2y ago·view thread
comment
Would be strange if they couldn't train ai on their work. Imo anyone should be able to
2y ago·view thread
comment
Good
2y ago·view thread
comment
What the hell? This is possible?
2y ago·view thread
comment
I agree with this. If there is close to zero cost for creativity and works of art then there does not need to be an incentive through copyright which gives a was designed to give a timed monopoly on w…
2y ago·view thread
comment
Isn't that the copyright maximalist clown suing ai companies?
2y ago·view thread
comment
>We expose a surprising failure of generalization in auto-regressive large language models (LLMs). If a model is trained on a sentence of the form “A is B”, it will not automatically generalize to …
2y ago·view thread
comment
Congrats to blinkDL
2y ago·view thread
comment
>Simultaneously, we're 10 years away from a world where staff "software engineers" go the way of human computers¹ Where's the source that this is going to happen?
2y ago·view thread
comment
Just pirate your collection
2y ago·view thread
comment
Has there been any voice cloning in any big commercial projects yet? It's mostly private people running models on their own machine/ doing it on 11 labs for fun. I don't see Hollywood d…
2y ago·view thread
comment
I doubt 1 to 1 cloning an actors voice without their consent to replace them is even a threat to actors as they have such an easy recourse. I see the future much more in actors that aren't modele…
2y ago·view thread
comment
Really appreciate what you're doing. Don't let those danish bottom feeders get you!
2y ago·view thread
comment
Looks like a complete nothingburger lol
2y ago·view thread
comment
Perfect. Now I finally don't have to click the links
2y ago·view thread
comment
Big fan of your work GGML friends
2y ago·view thread
comment
Yann bless
2y ago·view thread