Coming off a busy day filled with client calls (I am a full-time freelancer / maker), I had a tiny idea: could I get AI to paint scenes from classic books?
I decided to see if it's possible for me to build something small around this idea. I copied a block of text from David Copperfield, pasted it into DALL-E, and found out it didn't work: it needed a more cohesive prompt.
So I put the block on text into GPT-3, and then had it generate the image prompt for me. I then copy-pasted this prompt onto DALL-E, and it worked surprisingly well! (image)
I thought spinning this into a Twitter bot would be fun, and did just that, while tweeting simultaneously :)
While this is just a fun little experiment, and essentially all I did was glorified plumbing — I know nothing about AI or ML, and this is just me integrating with different APIs — I would be delighted to hear what you think!
Top comments (0)