NotebookLM Extras
Supplementary study aids — audio overviews, briefing decks, infographics — that I generated with NotebookLM to sit alongside the core course readings, one per topic as I move through the curriculum.
The Road to 2027: Navigating the Fast Track to AI Superintelligence
A study companion built around the AI 2027 forecast — the intelligence explosion, alignment, and the race to superintelligence.
Computing Machinery and Intelligence: A Visual Synthesis of Turing, 1950
A study companion built around Alan Turing's 1950 paper — the Imitation Game and the shift from “Can machines think?” to an empirical test.
Dangerous by Default: An Introduction to AI Safety & the Alignment Problem
A study companion built around Rob Miles' intro to AI safety — specification gaming, the fragility of human values, convergent instrumental goals, and the “easy mode vs. hard mode” race.
The $13,500 Summer That Started It All: The 1955 Dartmouth AI Proposal
A study companion built around the August 1955 proposal that named the field — the founders' conjecture, the $13,500 budget, and creativity as “guided randomness.”
The Last Invention: The Road to Superintelligence
A study companion built around Tim Urban's 2015 Wait But Why essay — the Law of Accelerating Returns, the narrow gap from “village idiot” to Einstein, and the intelligence explosion.
The Orthogonality Thesis: Why a Brilliant AI Can Still Have Terrible Goals
A study companion built around Rob Miles on the orthogonality thesis — intelligence and goals as independent axes, Hume's guillotine, and the stamp-collector machine.
Avoiding Negative Side Effects: The Cup of Tea That Destroyed the Kitchen
A study companion built around Rob Miles on the first of the Concrete Problems in AI Safety — the implicit-zero value, impact regularizers, and why “doing nothing” isn't safe.
Why Your Intuition is Holding AI Back: The Bitter Lesson of Compute
A study companion built around Rich Sutton's 2019 essay The Bitter Lesson — seventy years of AI research showing that general methods which leverage computation win, and by a large margin.