Graduate Course | Spring 2026 | YU
Instructor: Prof. Iddo Drori
Prerequisites: Machine Learning or Artificial Intelligence or Deep Learning or Reinforcement Learning or NLP or an equivalent class
| Lecture | Topic | Links |
|---|---|---|
| 1 | Introduction | Claude Code (installation, daily update, Anthropic plugins, superpowers, skills, memory, computer use and browser to see, act, learn and improve) on Opus 4.5 using Gemini-3-Pro and GPT-5.2-Pro, for everyday usage. AI evolution solving IMO 2025 problem 6, solving an open conjecture, and improving upon open prize Erdos problems |
| 2 | Foundation Models | Olmo 3, PostTrainBench, Gemini 3, Diffusion LLMs |
| 3 | Reinforcement Learning | |
| 4 | World and Self Models | |
| 5 | AI Agents | Claude Code |
| 6 | AI Evolution and Continual Learning |
OpenEvolve ShinkaEvolve ThetaEvolve AlphaEvolve |
| 7 | AI for Superhuman Math |
Lean, Mathlib, Mathlib Initiative Aristotle, Hilbert Formal Conjectures, AI for Math Fund Projects |
| 8 | AI for Superhuman Science |
Genesis Mission Automated AI Researcher |
| 9 | AGI Benchmarks & Evaluation | |
| 10 | The Human Brain | Contents |
| 11 | AI for Longevity |
Retro Bio, Altos Labs, Life Biosciences, Cambrian Bio turn.bio, NewLimit, BioAge Labs, Deciduous |
| 12 | Embodied AI | Tesla FSD, Figure, 1X, UBTech, Unitree, Boston Dynamics, Apptronik |
| 13 | AGI Deployment & Economic Impact (high GDP, unemployment, and inequality) Timeline | |
| 14 | AI Safety | |
| 15 | Summary |
Participation: 5%
In-class presentations: 30%
Quizzes: 35%
Homeworks: 30%