Kevin Madura

Qwen3.8-Max arrived with strong multimodal results, so I wanted to try something more concrete than visual Q&A: give it an image, ask for structured bounding boxes, and draw the answer back onto the pixels.

Satya Nadella recently published an excellent article on what a future firm looks like in an AI-driven economy.

… or how to process DataFrames with RLMs and DSPy Background: Recursive Language Models After experimenting with Recursive Language Models (RLMs) since their publication, I’ve been continually impressed by the powerful simplicity of the approach.

Before we begin: Of course this doesn’t replace a proper (human) audit or security testing process, nor should it.

Originally written September 2025 - a lot has changed between now and December when I’m posting this!

There’s been lots of discussion recently on DSPy and the GEPA optimizer.

DSPy is becoming increasingly popular (at least in my bubble on X) - and, in my opinion, for good reason!

Link to Paper A paper released in March 2025 by researchers at Model Evaluation & Threat Research (METR) evaluated the ability of AI models to complete long-horizon tasks.

Tools Used: bat ffmpeg llm* mlx_whisper pdftotext shot-scraper* uvx *Big thanks to @simonw for an unending flow of useful tools which makes life easier every day.

Large Language Models (LLMs) have burst into the conversation and have already proven incredibly powerful in accelerating all sorts of knowledge work.