A ChatGPT for Music Is Here. Inside Suno, the Start-up Changing Everything
Suno wants everyone to be able to produce their own pro-level songs — but what does that mean for artists? BY BRIAN HIATT for Rolling Stone.
We really need a different term than ‘AI’ to describe private, on-device machine learning that benefits individuals but (a) doesn’t violate anyone’s privacy, (b) doesn’t destroy the environment and, (c) doesn’t enrich smug Silicon Valley tech douchebros like Sam Altman.
Can we just stop using term #LLM is hallucinating, while it should be called “bullshting”. If you feed sht into it, sh*t will go out, same as with human education and we will never call it hallucinating for humans. Stop pretending it is some disease, it is fundamental flaw in design, bad use case and bad data. #ML
Thank you @RATPAC and everyone that came out tonight to talk about AI/ML in #hamradio. Lots of really interesting stories, achievements, questions, and answers. Video should be up soon and we will have more practical AI/ML sessions in the future.
Have you used #AI#ML in amateur radio? Would you be willing to talk about it? Let us know!
Allez, petit article qui va bien, tapé à l'arrache, mais qui peut vous intéresser. Comment j'ai utilisé une #IA, locale, pour générer de la data fictive.
Code fourni en bas de l'article. Et n'hésitez pas à réagir dans la section commentaire !
A conversation that keeps popping up in my mind since FOSDEM centers around open source projects and “AI,” and I still don’t know what I think. So let me share some thoughts here on the famously nuance-friendly Internet. 😜
During a chat w/folks from several open source organizations, someone suggested GNOME could attract funding by “sprinkling some AI on it.” Several folks laughed at the topical joke, but then realized it was in earnest. 🧵
My gut reaction—and judging by the room, the reaction of others—was “hell naw.” But then the person clarified, and I understood what they were actually saying. And I think it mostly jives with my thoughts around “AI.”
The suggestion, as best I can describe it, was that there are a lot of opportunities for genuinely useful, responsible, offline, machine-learning-powered improvements to a platform like GNOME.
For example: object recognition in the Image Viewer app to remove backgrounds; algorithmically improved camera quality in video calls; autocorrect! These are all areas that use ML algorithms on other platforms, and I don’t think that’s bad; you take a bunch of data, train an algorithm, then ship that in the OS/GNOME/etc. to be genuinely helpful.
First, we can get on board with that, right? Personally, I don’t consider that “AI,” even.
So… how does that attract funding so that we can invest in the platform as a whole, making everything (not just these hypothetical ML-powered improvements) better—funding that could be put towards continuing to improve accessibility, developer APIs, research and user testing, and features that our users care about?
Well, you kind of have to pitch it as AI. 😬️ Let me explain.
Like it or not—and I absolutely hate it—“AI” is where the hype is now. I don’t want to play into it. I think “generative AI” is extraordinarily overblown, problematic, and burning an immeasurable amount of resources. So let’s not do that.
But if you could secure a grant to “integrate helpful, privacy-respecting AI features” into GNOME… wouldn’t you? To be clear, this would include the features I mentioned earlier—not generative AI.
So I guess my question is (sorry this is so rambly, nuance is hard!!)… how would you as an ethical open source project with diverse interests and a community of users with strong feelings around “AI” approach this?
I almost feel like we need to distinguish the current generative AI hype from ML-powered features when it comes to public communication—but also lean into “AI” when looking at funding. And that distinction is… hard.
If you work with spectra or multivariate regression and don't want to reinvent the wheel, check it out. If it doesn't do what you need it to do, let me know and we can add capabilities to make it work for you! #python#spectroscopy#lpsc2024#data#OpenSource#DataAnalysis#ML