[453qtgreq]

Parent commenter here. No-one has picked up on the last of my points, which I think is the most crucial. Once we're all hooked on the "productivity gains" of widely using GPT, what will we do once Microsoft start bundling it into subscription packages? Just pay up, I suppose.

[CuriouslyC]

There are a variety of GPT models with different numbers of parameters. I can run GPT-J locally right now, and as model architectures progress and the killer app for video cards goes from games/mining to AI, we will see a combination of increased card VRAM and memory optimized models that will make running this stuff locally feasible. Training is expensive, but inference is usually fairly cheap, you just need to be able to load the model in VRAM. Thankfully there are open source groups and communities working together to train and fine tune models of this sort so we won't be completely reliant on commercial entities.

[jason-phillips]

You appear to me to be stuck in the academic world of arguing concepts that are detached from how the simplest of things work in the real world.

> [W]hat will we do once Microsoft start bundling [GPT] into subscription packages?

Pray tell, how will Microsoft do that, exactly? And assuming there is some licensing or subscription model for the use of GPT, what is wrong, exactly, with that?

[453qtgreq]

I don't understand what you mean by your first sentence. Copilot-enhanced VS Code is becoming very popular. Copilot uses GPT3, which is exclusively licensed by Microsoft from OpenAI. I'm very much talking about the real world, and the frighening ease and speed which which MS is taking over the way we write software.

>Pray tell, how will Microsoft do that, exactly?

How will Microsoft make people pay for tools that they currently distribute freely? You don't have to look very far to find examples.