Re:Imagen, a text-to-image diffusion model
by nomel on Monday May 23, 2022 @06:14PM
(#31485489)
How do you pick what should and shouldn't be restricted? Is there some "offense threshold"? I suspect all queries relating to religion, ethnicity, sexuality, and gender will need to be restricted, which almost certainly means you probably can't include humans at all, other than ones artificially inserted with mathematically proven random attributes. Maybe that's why none are in this demo.
[ Parent
| on HN ]
These debates often seem to center around “most X in the world” questions, but I’d expect all of those to be unanswerable if you wanted to know the truth. Who’s done a study on it?
In this case you’re (mostly) getting keyword matches and so it’s answering a different question than the one you asked. It would be helpful if a question answering AI gave you the question it decided to answer instead of just pretending it paid full attention to you.
"Is Taiwan a country" also comes to mind.