I completely forgot that Holden was part of that commune of, what, eight mothers?

It’s a shame, because ComfyUI can be so much more than just image generation. And just because there’s a lot of string processing for LLMs doesn’t mean that it isn’t important to capture in an I/O interface, especially when it comes to preserving chat history. Save data, load data, ask new questions, etc.

ChatGPT is pretty damn powerful, I’ll admit. But, all of its components need to be localized, especially since something like a Mixture of Experts model could be split down to base models and loaded/unloaded as necessary.

ComfyUI is just a bunch of Python code tied into I/O nodes. I’m surprised there isn’t a good set of nodes for SGLang yet.

This seems like the kind of thing that should be bookmarked and thrown in corporations’ faces at every moment.

Have you used any good ComfyUI workflows specifically for chat LLMs?

Yeah, no. Not buying that Elon is trans.

My hope is that the “AI” craze culminates in a race to the bottom where we end up in a less terrible state: local models on people’s phones, reaching out to reputable websites for queries and redirection.

We’re already heading there. But, it’s not going to happen by sitting on your hands and waiting for the billionaires to hand you these local models on a silver platter. You honestly believe the overlords that own your phone will give you shit for free? They want you hooked on subscriptions, that send all of your personal data and social security numbers to their huge databases, until the day you die. And then they’ll sell that data to your children and your grandchildren just to make even more profit.

You have to take it. You have to find it yourself. You found Lemmy. Good. So, go find other shit. Discover open source. Discover piracy. Discover Linux. Stay on top of it.

Google just killed uBlock Origin, but I’m using Firefox, because the writing was on the wall at least a year ago.

I stared at it for 10 seconds, which is double the advertised time it’s expected to explode. I want a refund!

Israel isn’t Judaism and Judaism isn’t Israel! Mixing that up is problematic because you’re confounding anti-zionism (a settler colonial state project committing genocide) with anti-semitism.

Mixing that up is intentional, so that you can be accused of being antisemitic by not supporting Israel. Politicians love to do this.

Lies. I see no self-destruction in action here.

Is it also down-to-up? Because it doesn’t make sense otherwise.

AI existed before LLMs. People called NPCs and enemy character patterns “AI”.

Trying to take off the arm covers with her mouth?

The Federal Bureau of Investigation (FBI) has seized the domain names of several gaming-related pirate sites, including NSW2U.com, which has been Nintendo’s nemesis for years.

Ahhh, there’s the real reason. Nintendo put pressure on the FBI, because Nintendo.

Or not buying any new Ubisoft game.

Same company acquired two very similar apps.

One required a $14/year subscription, the other $90/year

So, why are you using a subscription at all, and not a free and open source app?

You know what Blender is? It’s the industry standard. Why is it the industry standard? Because it’s free, open-source GPL, it can never be taken away from you, it can never be “acquired”, everybody’s using it, and everybody is contributing towards improving it. Nobody is bitching about how Blender is extracting millions of dollars from its users and reducing its feature set at the same time. (*cough cough* Photoshop *cough*)

Free and open source software. Every. Single. Fucking. Time. Fuck subscriptions.

Not even mods can do this in Reddit, though. Seems like an unreasonable restriction.

If only there was a system that logged every moderator action into a public page that everybody can see. Maybe we can call it “modlog” or something like that.

So, UN experts then. Seems more trustworthy than Israeli experts.

That’s not true. There are other ways of influencing the numbers that tools use. Most of them have their own internal voting systems, so that humans can give feedback to directly influence the LLM.

Diffusion models have LoRAs, and the models themselves can be trained on top of the base model.

Your co-worker is bad at his job, and doesn’t understand programming.

LLMs are cool tech, but I’m gonna code review everything, whether it comes from a human or not.