Keynote: Red, Blue, and Purple AI
summary
Talk claims practical, ready-to-run artifacts: prompt methodology that yields 30–35% improvement over single‑shot prompts, system prompts tuned for ChatGPT‑4, Claude, Llama3 and Azure OpenAI, and a toolkit of working bots — subdomain doctor, an acquisitions finder that surfaces companies Crunchbase misses, a system bot that scaffolds new security GPTs, XSS mutation suites (~40 mutations), nuclei/YARA/OSQuery generators, and Splunk/ELK playbooks; demos and real engagements report successful detection, enumeration, and automation.
AI-written summary · about
chapters (9)
- 0:00Generative AI in security
- 1:04AI from sci-fi to tools
- 10:18How LLMs work
- 17:21Choosing AI models
- 23:21AI bots for security
- 38:43Red team AI automation
- 49:11Blue team AI defense
- 58:23Purple team AI tools
- 1:00:51Q&A
related talks
47mThe Art of Prompt Engineering: Best Practices for Effective AI Interactions
01h 08mIn Prompts We Trust
22mAI, Make It So - A Dive into the Universe of Prompt Engineering
19mKeynote: Towards Building Safe and Secure Agentic AI
02h 01mPrompt Engineering and AI Red Teaming
58mKeynote: The past, present, and future of AI for application developers
44mThe Science of Effective Prompting
24mKeynote - The AI Playbook For Companies