# History, Westworld, and Writing Heuristics for Thinking About AI *By Recommended Reading from Tech Founders • June 13, 2026* Today’s strongest organic recommendations were compact frameworks: Satya Nadella’s history pick on technology, markets, and democracy; Sahil Lavingia’s Westworld heuristic for agent loops; Brian Armstrong’s Paul Graham prompt for concise writing; and Keith Rabois’s thread on frontier AI as strategic capability. Sarah Go on "untrainable parts" and Paul Graham’s case for autobiographies rounded out the people-side of the list. ## What stood out Today's cleanest recommendations were **working frameworks**, not just links to save. Satya Nadella reached for history to explain the balance between technology, markets, and democracy [^1]; Sahil Lavingia used *Westworld* as a shortcut for understanding agent loops [^2]; Brian Armstrong shared a Paul Graham post as a prompt for more concise AI-assisted writing [^3][^4]; and Keith Rabois pointed readers to a thread he said was worth reading very carefully because it framed frontier AI as controlled strategic capability [^5][^6]. A second theme was human capability: Nadella highlighted Sarah Go on the "untrainable parts" of people and organizations, while Graham suggested autobiographies as a better parenting resource than most parenting books [^1][^7]. ## Most compelling recommendation ### *Parallel Paths to Prosperity* - **Content type:** Book [^1] - **Author/creator:** Joel Mokyr and co-authors [^1] - **Link/URL:** Not provided in notes - **Who recommended it:** Satya Nadella [^1] - **Key takeaway:** Nadella said the book describes a thousand-year history of Western growth and the virtuous cycle among technological revolutions, markets, and democracy, each acting as a check on the others [^1] - **Why it matters:** This was the strongest pick today because Nadella did more than name the book; he extracted a concrete framework and applied it to how society should think about the current age [^1] > "the west in particular got three things into a virtuous cycle... technological revolutions and markets and democracy all both acting as a check on the other" [^1] [![Satya Nadella ơn A.I. Jobs: Humans Will Do the ‘Glue Work’](https://img.youtube.com/vi/zqEZyHkXgh0/hqdefault.jpg)](https://youtube.com/watch?v=zqEZyHkXgh0&t=1498) *Satya Nadella ơn A.I. Jobs: Humans Will Do the ‘Glue Work’ (24:58)* ## Compact frameworks for the AI moment ### *Westworld* (first episode) - **Content type:** TV episode [^2] - **Author/creator:** Not provided in notes - **Link/URL:** Not provided in notes - **Who recommended it:** Sahil Lavingia [^2] - **Key takeaway:** Lavingia said that if you do not understand agent loops, you should watch the first episode of *Westworld* [^2] - **Why it matters:** It was the most compressed recommendation in the set: one episode offered as a mental model for how agent loops work [^2] ### Paul Graham post on concise, "unsummarizable" writing - **Content type:** X post [^3] - **Author/creator:** Paul Graham [^3] - **Link/URL:** [https://x.com/paulg/status/2062891972042637573](https://x.com/paulg/status/2062891972042637573) [^4] - **Who recommended it:** Brian Armstrong [^3] - **Key takeaway:** Armstrong said Graham's quote is a good prompt for an agent to make writing more concise without stripping out the interesting ideas [^3][^4] - **Why it matters:** It is a practical heuristic already being used in real writing workflows, not a generic preference for brevity [^3][^4] > "strive to make my writing unsummarizable, in the sense that it has so little fluff left in it that if you take any words out, as summaries by definition do, you lose a lot of interesting ideas" [^3] ### Thread on frontier AI as controlled strategic capability - **Content type:** X thread [^5] - **Author/creator:** Linked X account: `_the_prophet__` [^5] - **Link/URL:** [https://x.com/_the_prophet__/status/2065613953288597855](https://x.com/_the_prophet__/status/2065613953288597855) [^5] - **Who recommended it:** Keith Rabois [^5] - **Key takeaway:** Rabois said it was worth reading very carefully; the linked post called the shift a "monster signal" and described it as the moment frontier AI stops being treated like software and starts being treated like controlled strategic capability [^5][^6] - **Why it matters:** It was the clearest recommendation today about a change in how frontier AI is being framed [^6] ## Human capability and indirect learning ### Sarah Go on "untrainable parts" - **Content type:** Blog [^1] - **Author/creator:** Sarah Go [^1] - **Link/URL:** Not provided in notes - **Who recommended it:** Satya Nadella [^1] - **Key takeaway:** Nadella said Go's post asks what the untrainable parts are in people and organizations, and he connected that to human agency, ambition, and the "glue work" people do [^1] - **Why it matters:** It was the clearest people-centered recommendation in today's set, emphasizing capabilities Nadella said should not be counted out [^1] ### Autobiographies (especially the early sections) - **Content type:** Book category [^7] - **Author/creator:** Various - **Link/URL:** Not provided in notes - **Who recommended it:** Paul Graham [^7] - **Key takeaway:** Graham said early sections of autobiographies are often implicitly about parenting, making them a better recommendation than most books explicitly about parenting [^7] - **Why it matters:** It is a distinctive reminder that some of the best practical learning comes indirectly through life stories rather than advice manuals [^7] --- ### Sources [^1]: [Satya Nadella ơn A.I. Jobs: Humans Will Do the ‘Glue Work’](https://www.youtube.com/watch?v=zqEZyHkXgh0) [^2]: [𝕏 post by @shl](https://x.com/shl/status/2065573422634053991) [^3]: [𝕏 post by @brian_armstrong](https://x.com/brian_armstrong/status/2065593892721025026) [^4]: [𝕏 post by @dorvonlevi](https://x.com/dorvonlevi/status/2065580214227243244) [^5]: [𝕏 post by @rabois](https://x.com/rabois/status/2065615589918281998) [^6]: [𝕏 post by @_The_Prophet__](https://x.com/_The_Prophet__/status/2065613953288597855) [^7]: [𝕏 post by @paulg](https://x.com/paulg/status/2065369429416825090)