Not heard about NotebookLM?
No worries. You are in good company...
I had not heard about it either, before I came across the NotebookLM: Your digital Hype man! video on Youtube, the very same day it was released.
So I had a go at it, letting it loose on this site (raku-musings.org).
The up front result was a short textual summary, in my native Norwegian language:
Arne Sommers "Raku Musings" er en samling av artikler som utforsker ulike aspekter ved Raku-programmeringsspråket. Mange innlegg demonstrerer praktisk bruk av Raku for å løse spesifikke problemer, som generering av navn eller anagrammer. I tillegg presenteres oppdateringer om Raku-prosjekter og diskusjoner rundt språket selv. Nettstedet inneholder også en omfattende liste over ukentlige programmeringsutfordringer som kan løses med Raku.
NotebokLM is a Google Offering, and they obviously know that I am Norwegian. The impressive part about the summary is that the musings site is 100% in English (except for some gibberish), so they have translated the summary automatically. Without any obvious errors.
This is the result of feeding the summary into Google Translate, sort of «back to Mama»:
Arne Sommer's "Raku Musings" is a collection of articles that explore various aspects of the Raku programming language. Many posts demonstrate practical uses of Raku to solve specific problems, such as generating names or anagrams. In addition, updates on Raku projects and discussions surrounding the language itself are presented. The site also contains an extensive list of weekly programming challenges that can be solved with Raku.
A pretty good translation...
Albeit rather shallow on details (as is the Norwegian version, of course)...
The really interesting part of NotebookLM is the generation of a podcast, a two person discussion of the content. In English.
Here is the 12 minute podcast.
It is pretty good, until they get to my Time Balls with Raku article (which was the newest article when I did this), where they dropped the ball (pun intended) and entirely missed the point of the article.
Listening to the podcast several times made it obvious that the conversation format made a lot of the vagueness less obvious. A pretty slick way of obfuscating the shortcomings of NotebookLM.
Further reading: