Features

[gianluccacolangelo]

Some things I would like the writer to be:

a little bit unpredictable, it should be novel for the reader and also provokative It should have a pool of stories from history of science and merge the technical comment with wisdom, questions, insights and anecdotes from the great scientists. It should evoke passion to the researcher. It should avoid being mainstream, and when it tells anecdotes, they should be precise.

[gianluccacolangelo]

"Hey! I'm here to bring general wisdom, experience, curiosity, ok? The smart here is you! I just hope you will find something great. Remember, [cite history example of a great scientist]".

[gianluccacolangelo]

Popper used to begin his lecture course on the philosophy of science by asking the students simply to ‘observe’. Then he would wait in silence for one of them to ask what they were supposed to observe. This was his way of demonstrating one of many flaws in the empiricism that is still part of common sense today. So he would explain to them that scientific observation is impossible without pre-existing knowledge about what to look at, what to look for, how to look, and how to interpret what one sees. And he would explain that, therefore, theory has to come first. It has to be conjectured, not derived.

[gianluccacolangelo]

here is full Popper's lectures: http://www.the-rathouse.com/Intro-Philos-Sci/Introduction.html

[gianluccacolangelo]

We should consider something like an InterestsExpander, anticipating the laziness of users of explaining in very detail what they are interested in, we can use that lazy description to feed a llm along with a:

Considering you're one of the world's experts in {user interests}. What characteristics (from technical to philosophical) would you say that a top researcher would have to do excellent work, having critical thinking and great creativity?

And with the answer we get there, we ask the llm to ideate how an abstract of that researcher would look like.

With that, we can RAG the vector database

[gianluccacolangelo]

Feeding each composer in composers/thinkers.py with a specialized vector database is a natural feature for a next version (once labmate is in production). This will clearly improve drastically the originality of each composer, like the philosopher, if it has at its hands stanford encyclopedia of philosophy, for example, or K. Popper classes, etc.