At the end of October, I was in Paris, France for the M3AAWG 68th General Meeting, held October 26 to 29, 2026.
Together with Gabriel Loiseau, I gave a training session:
🇬🇧 LLM Optimization in Practice: Better Results Without Fine-Tuning or Wasting Tokens
Sometimes, you rely on models you cannot fine-tune. One way to get better results is then to improve the instructions you send to the LLM, while keeping the cost low by using as few tokens as possible.
As lead of the AI SIG, I also moderated several sessions. These meetings are always a highlight of my year: the chance to swap ideas with people who fight abuse every day is hard to find anywhere else.
Thanks to everyone who attended, and see you at the next one!

