The fact that we're starting to get access to these huge pretrained models offline, for free, and in our local machines, is insane.
Whisper let's us transcribe and translate, Stable Diffusion generate images, and GPT-like models generate text.
There's a lot more to come.
We're just getting started.