artninja1988

Agreed. Honestly gpt4o is a massive disappointment. They made it multimodal, but it seems as dumb as ever reasoning wise. I think it's become pretty clear that we've reached diminishing retu…

Wrong link?

I think there's a spectrum among employees to how much they believe in the xrisk/ lesswrong stories. The people on the board were full on effective altruists. If I had to guess he's com…

Anthropic desperately needs a product to differentiate itself it seems to me. Basic LLM foundation Models are getting commodotized fast and it seems there are diminishing returns setting in, despite w…

https://archive.ph/is57b

11b model outperforms 8b model, news at 11

> Giving social networks complete control over the collection and use of public web data "risks the possible creation of information monopolies that would disserve the public interest," …

Yeah, with these generators you can input your own lyrics for the model to sing and make a song with

Is anyone working on an opening source music model now that stability is dead?

Elevenlabs rap. Man, we are living in the future... https://x.com/flavioschneide/status/1788654503790628935 …

Seems fair enough. Why was the factory allowed to be built there if it threatens the drinking water?

New codex version or what do you have in mind? Something like Devin that actually works?

>infringed on copyright It still isn't clear if training on copyrighted data is infringement or not. Please stop spreading misinformation

Tldr. is that he's vaguely arguing for neuro symbolic ai without proposing anything specific that's implementable let alone building it himself. It's just snarky pointing out flaws and …

These models will exist one way or another. Even the most IP concerned lawmakers in the deepest pockets of the RIAA and whatnot. It's just that some want there to be monopoly rents payed to thems…

Grim. This is just confirmation that the future of llms will be a walled garden of big tech and big rights holders. There's no open ecosystem if content needs to be licensed at prohibitive prices…

Why do you say he's a troll?

Is this the same news outlet that got on civit.ai and stabilities case for being "complicit" in unsavory generated pornography? Odd that there's a different slant to this story

>using significant sections of such works for training generative AI software is copyright infringement Does the author of this article have any sources for this?

(29. January 2024)

>Meaning legal system will have to bend towards reality and copyright is dead in LLM context and thus many other contexts as well. I certainly hope so. There are already a few text and data mining …

http://archive.today/fS9fR

Dupe: https://news.ycombinator.com/item?id=40071858

Yeah, it's not going to be SOTA but existing models are already good enough for a reasonable portion of use cases. Even if they lag behind it's important to have free models.

I mean it's in the crown sourcing range for diffusion models and going to get cheaper. It currently costs 50k to retrain stable diffusion 2 ( https://www.databricks.com/blog/s…

The whole differentiator between firefly, the adobe image generator, is that it's trained on their licensed stock image library and commercially safe. Besides that, even openai is licensing some …

Great work and I'm excited to see more competition in this space.The vocals do seem more impressive / less wooden than Suno. Would really love to see a future of more controlability and sepa…

Getty wants every model developer to pay Getty. Big news