Things I have read so far (2025)

Virtual Machinations: Using Large Language Models as Neural Computers: LLMs can function not only as databases, but also as dynamic, end-user programmable neural computers.

An incredible read on how LLM-based software systems can be the next generation of computers. Belief is that the LLMs are much more than a database of information and can be thought as branch predictor on top of the instruction set (tools).

It also draws similarities between today’s LLM-based system components and traditional computer architecture and argues the issues with such systems (retrieval and injection of relevant context) can be solved with the solutions traditionally embedded in hardware (paging, named operands, etc.)

Getting AI to write good SQL: Text-to-SQL techniques explained

Assignment 3: Advanced Shell Scripting and Workflow Automation