Limited number of people every week

Ep 1&2 - "AI Intution" Notes by AI

• Ice Breakers: Started with some fun introductions.
• Knowledge Map Overview:
  • Highlighted critical resources: Gemini 1.5, Tori (GPT-4 preview unlimited), ChatGPT-4, Claude 3, and other AI and local tools like Comfy UI, Otter.AI, Goblin Tools, etc.
• Predictions:
  • Discussed the future of open-source AI, potential for non-biased training sets, and AI surpassing government compute capabilities.
  • Explored legal challenges, including anti-impersonation, copyright issues, and the concept of AI creating fake human identities.
• This Week's Trends:
  • Talked about AI's impact on the labor market, communication enhancements, the importance of online learning, and the AI arms race.
  • Considered how AI might change traditional fields like law by finding loopholes and contradictions.
• Future of AI:
  • Emphasized the need for understanding AI tools for success.
  • Discussed the potential for AI to revolutionize industries and traditional systems.
• In-depth Topic:
  • https://www.youtube.com/watch?v=kCc8FmEb1nY (Lets build GPT from scratch) by Andrej Karpathy

Let's be clear about goals

• Expectation is at least 2 hours, 2 x week.
• Optional Group study sessions weekly
  • Tuesday 6-8pm ET
  • Thursday 6pm-8pm ET
• Membership into subsequent seasons is based on commitment

Topics we'll review include:

• Building intuition behind AI
• Basic AI alignment + different attack vectors (worms, injection, trojans, poison pills, jailbreaks)
• Review on AI tools for creating AI images + videos (loras, midjourney, comfy UI, A111, llama, Stablediffusion)
• State of Art Techniques (Byteformers, 4d Gaussian Splatting, SIMA, Neuro-evolution)
• Strategic analysis of AI actors (China, TSMC, US Gov DOD, Google, Msft, OpenAI, Meta, Apple, Nvidia, ASML, ZEISS)
• Discussions on AI-proof jobs (and their respective time frames)