How much VRAM do you need to run an LLM locally?
Work out the GPU memory a local LLM needs from its parameter count, quantisation and context length, with worked examples from official docs.
Read articleTheAILAB blog
Notes from the lab.
We write about the things we build: automating everyday business work, running AI on your own hardware, and turning trading rules into automated systems. When we state a fact, we link to where it came from.
Why your backtest looks better than your live trading results
Overfitting, look-ahead bias, survivorship bias and missing costs explain why most backtests disappoint in live trading. Here's how to catch each one.
Read articlePrompt injection: the security risk every AI automation must plan for
Prompt injection lets text hidden in emails, documents or web pages hijack an AI assistant. What it is, why there's no full fix, and how to design around it.
Read article