FaceDeer

joined 1 year ago
[–] FaceDeer@fedia.io 2 points 14 hours ago (2 children)

There's also Udio.com and Producer.ai out there, and possibly some others - music generation is becoming fairly widespread. I didn't mention any of this in my list of recommendations though because OP specifically asked for LLMs. :)

[–] FaceDeer@fedia.io 2 points 16 hours ago

I've currently got a fancy Python script using that model to scan through 15 years worth of transcripts of audio recordings figuring out what the context and subject matter of the recordings are, so that I can sort and search them. Previously they were just a giant pile of audio files.

I used WhisperX to transcribe them. The hardest part was getting pip to install WhisperX correctly, everything else has been just routine Python coding.

[–] FaceDeer@fedia.io 2 points 16 hours ago

And if you're both concerned about bugs and also don't have the time to verify it all manually, you can specify that you want the code to have plenty of sanity-checking and error logging functionality in it so that if something goes wrong you'll know immediately.

AI codegen is great for creating unit tests because human programmers never even bother to create unit tests most of the time.

[–] FaceDeer@fedia.io 7 points 16 hours ago (2 children)

https://notebooklm.google.com/ is really handy for various things, you can throw a bunch of documents into it and then ask questions and chat interactively about their contents. I've got a notebook for a roleplaying campaign I'm running where I've thrown the various sourcebook PDFs, as well as the "setting bible" for my homebrew campaign, and even transcripts of the actual sessions. I can ask it what happened in previous episodes that I might have forgotten, or to come up with stats for whatever monster I might need off the cuff, or questions about how the rules work.

Copilot has been a fantastic programming buddy. For those going a little more in depth who don't want to spring for a full blown GitHub Copilot subscription and Visual Studio integration, there's https://voideditor.com/ - I've hooked it up to the free Gemini APIs and it works great, though it runs out of tokens pretty quickly if you use it heavily.

[–] FaceDeer@fedia.io 3 points 1 day ago

Of course not. But it changes the economics that causes things like data centers to be built there.

[–] FaceDeer@fedia.io 3 points 1 day ago

There are many other inefficient uses of water much bigger that these data centers. Texas has a major agriculture sector despite being basically a desert. People love to have green lawns. And a quarter of Texas is currently in a drought.

My point is that if you shut these data centers down right this instant the needle isn't going to budge much.

[–] FaceDeer@fedia.io 5 points 1 day ago (3 children)

So charge them more for the water. The data center builders are making rational decisions based on their costs.

[–] FaceDeer@fedia.io 2 points 1 day ago (2 children)

Perhaps there are reasons beyond just these data centers why Texas has a water shortage.

[–] FaceDeer@fedia.io 1 points 1 day ago

Was this concrete batch plant being built by one of the richest companies in the world?

If this was the case for literally every project, then every project would cause cities to lose money and I have no idea how cities could still exist.

[–] FaceDeer@fedia.io 4 points 1 day ago (1 children)

Palestine Action spray-painted some stuff. They didn't "destroy Israeli weapons factories", that's rather hyperbolic.

[–] FaceDeer@fedia.io 1 points 1 day ago (2 children)

It's not just jobs as in the physical bodies that are going in to work at the place, it's also the taxes and the net expansion of the local economy (eg, the people working there will be spending money at local restaurants and businesses, the facility will be paying for utilities, and so forth). It's a complicated thing to evaluate.

[–] FaceDeer@fedia.io 1 points 1 day ago (1 children)

There's lots of factors to consider beyond just water. Cost of power, cost of construction and staff, access to internet, proximity to demand for low-latency access, and so forth.

view more: next ›