Hacker News new | ask | show | jobs
by alphabetatheta 1178 days ago
I wonder how much we can extend this to image or video labelling tasks? this could really jeopardize the business moats of lot of data labelling startups and services.
1 comments

You can extend it to data labelling tasks and it will do pretty well.

https://imgur.com/a/3XfOBq1

This is stable diffusion processing a small, blurry image on my PC and correctly identifying it. It did take 7 minutes but my 1 GPU is no match for the ~250M USD worth of GPUs owned by OpenAI.