Things I Read

My rough bet having looked at technologies over time is that what is more likely to happen in the world of LLMs is that there is going to be an increase in value capture by fewer engineers than before, but that demand for engineers overall is going to go up. I view this similarly to the way that automated trading allowed market makers to increasingly capture more and more value as automation kept coming in, despite the cost of each trade going down, or the salaries of top football players continues to go up despite the game being more accessible than ever before. What happens is that overall demand goes up, and more of that demand is captured by a very skilled few.

I chose these specifically because:

1. Understanding what is actually happening at the lowest level of abstraction is going to be important.
2. Cool new technology, need to see what's happening under the hood
3. A classic

In order:

1. I've recently been thinking a lot about what the hell is going on under the hood of maths. One of my favorite quotes since I first read it about a decade ago:

   "It must have required many ages to arrive at the startling notion that a brace of pheasants and two days are elements of the same type."

   It's taken me some time admittedly to fully internalize the implications, but I'm trying to make some headway in understanding what is going on at the lowest level.

2. Connected to the books above, I'm doing a rough bottom-to-top, top-to-bottom approach for maths and computer science, and I think a good place to start is where you could, for most intents and purposes, say the whole thing kicks off in the past century, with Turing's paper on the Entscheidungs problem.
3. As I've been brushing up on my linear algebra for LLMs, I've been struck once again, by how little I actually deeply understood about what all of it meant. Sure you can do the calculations and hand-wave about concepts, but there is quite a lot of mathematical depth that is missed. This is the first step into getting a little more understanding of linear algebra in general.