A talk at The Hackers Meetup Ahmedabad on AI agents. It covered how people test and attack agents today and five open-source tools for doing it, and laid out six open problems in this domain.
- When
- 25 Oct 2026
- Where
- WeekMate, Ahmedabad
Loading…
Reading circle · Date to be announced
AhmedabadNovember 2026 · date TBA
Share this event
Our first reading circle works through Ngo, Chan & Mindermann (2022) — probably the best single overview of why standard training methods could produce AI systems that pursue goals their developers didn't intend. The paper is circulated a week before; the session is a short summary followed by open discussion.
About the organiser
SAGA is a research, discussion and study community on AI safety and alignment, based in Ahmedabad, India. We run reading circles, talks and meetups, free and open to everyone, and often speak at other groups' events too. Stay connected to hear about the next one.
A talk at The Hackers Meetup Ahmedabad on AI agents. It covered how people test and attack agents today and five open-source tools for doing it, and laid out six open problems in this domain.
A talk at NULL Ahmedabad's monthly security meetup that introduced AI alignment from the ground up: how today's AI systems are made, why that makes it hard to know what they have actually learnt to want, what researchers are doing about it, and how people in the room could help.
Sign up and we'll let you know about upcoming sessions, with pre-reads sent a week ahead.
