Language

 This morning I had a long chat with Gemini about language learning for adults. The similarity with my findings in chess learning is striking:

• Understanding. 95% of the learning should be based on understanding
• Shadowing. Only lately I learned to appreciate the learning based on a close observation of titled commentators who are playing or watching other players play.
• Frequency of occurrence. A long term pet subject of me. Focus on what happens in each and every game. Preferably every few moves.
• Spaced repetition. A necessary hack of the forgetting brain.
• Interest in certain aspects of the game. Only lately I am starting to develop a specific interest in the Vukovic gap of the game. Hence I put endgames on the backburner.

Prof. Elan Barenholz has an interesting hypothesis about how certain parts of our brain have a lot in common with LLM's (Large Language Models). As a child, our mind is filled with all kinds of knowledge. That starts before we can even speak. That is why western people have western biases (hallucinations) and Hindus have Hindu biases. That is why LLM's perform so well. LLM's don't think, they are just good at paroting. Just like we most of the time.

I'm sure we will obtain a few insights that are usable for chess, some day. Stay tuned.