Have you tried NotebookLM from Google already?

It has a few fantastic tools, but its Deep dive conversation generator absolutely blew my mind.

The Deep dive conversation generator creates realistic podcast conversations from the provided sources.

The generated conversations are engaging, funny, and they augment the provided source in unexpectedly good ways.

NotebookLM is one of the best AI tools I’ve ever tried.

📼 Here’s one example conversation generated from my paper Ants Can Play Prisoner’s Dilemma.

After 4 years, there’s a new version of PySwip. I’ve released v0.3.0.

PySwip is an open source library that enables running Prolog queries in Python via a foreign language interface to SWI-Prolog. It has no dependencies, and it works wherever SWI-Prolog runs, Linux, Windows, MacOS, OpenBSD, and elsewhere.

Here are some of the highlights from the change log:

• SWI-Prolog 9.x is supported.
• Dictionary support is addded,
• Improved list representations, unicode support and multiple threading usage,
• Refactored SWI-Prolog discovery.

If you have a recent version of SWI-Prolog installed, the following is sufficient to install PySwip in most cases: