Prompt Engineering and AI Red Teaming
summary
Creator of Learn Prompting and HackAPrompt describes concrete results: an open dataset of 600,000 prompts, a 76‑page Prompt Report surveying 1,500+ papers and cataloguing 200+ prompting techniques, and a training reach of ~3 million people. Empirical claims: chain of thought plus few‑shot gives top MMLU gains, role prompts fail on accuracy, 20 hours of manual tuning is outperformed by automated optimizers, and ~40% of successful jailbreaks transfer across models — prompt injection defenses are not fully patchable.
AI-written summary · about
chapters (7)
- 0:00Introduction and background
- 17:12Prompt engineering fundamentals
- 25:47Advanced prompt engineering techniques
- 1:09:43AI red teaming overview
- 1:22:01Prompt injection and jailbreaking
- 1:38:36AI security challenges
- 2:00:33Live competition and Q&A
related talks
13mPrompt Engineering Simplified
16mPrompt Engineering for Testers: Types and Applications
30mIs Prompt Engineering Dead? How Auto-Optimization is Changing the Game
44mSupercharge Your Productivity Mastering Prompt Engineering
01h 04mIn Prompts We Trust
47mThe Art of Prompt Engineering: Best Practices for Effective AI Interactions
30mIs Prompt Engineering Dead?
29mSupercharge Your Productivity: Mastering Prompt Engineering