AI tools for detecting image manipulation are coming into their own.
"Screening papers with the use of ImageTwin.ai increased the number of inappropriate duplications detected, with 41 of the 115 being missed during the manual screen and subsequently detected with the aid of the software." - Sholto David
This request might seem strange, but let me try and explain. Since we renewed our work on https://groupd.co/ a few weeks ago, I thought it would be a great idea to demonstrate how our ML models work with a small public demo.
I am training a "positive news" model which will filter out the news noise, and hunt for good-vibes-only news and articles.
It would be great, if you, folks send me a couple of good good-vibes-only favorites of yours.
Fascinating as they are, AI assistants are not works of logic; they are works of words. They have become incredibly good at producing text that looks right. For many applications that is enough, not for programming though.
I was sceptical regarding the DiagramGPT tool, but I made some experiments - free text, docker compose file, source code in multiple program languages. And it just works. And it has an API, making generating living documentation even simpler.