Curiosity as an intellectual virtue is relatively rare, especially among tenured academics; Cowan attributes his own openness to change partly to heritable traits (estimated ~60% heritable) and partly to his own choices/life experience.

When balancing openness against closure for economic competition, policymakers face an irreducible uncertainty: it is unclear whether it is better for America to attract a competitor's talent and make them dependent on US hardware networks, or to cut off all access and allow competitors to develop fully independent systems.

Cowan trained a companion AI model on his book 'GOAT' using consultant help; the model allows readers to query the book's content directly rather than reading linearly, functioning as an AI teaching assistant.

The distinction between innovation and imitation is Hegelian (dialectical) rather than oppositional: great innovators typically build on imitation and copying; claiming innovation happened in a vacuum while imitation is derivative misses how creative progress actually works.

The Great Stagnation (claimed to have run from ~1973 to 2020) has ended, marked by breakthrough innovations in mRNA vaccine development (which could be designed in one day and deployed in under a year, beating the 4-year expert forecast) and continued progress in AI and biomedicine.

The Great Stagnation was a symptom on the application layer only; innovation trends (Moore's Law, Transformer architecture, AI breakthrough at Deep Mind, internet scaling) continued beneath the surface, suggesting culture was 'fine' and we just needed to wait for applications to emerge.

People who have grown up in moving history (like Jonathan from China) are actually envious of the stability Western-born people take for granted; the 'grass is greener' effect applies to both directions—stable societies don't realize their advantage, and dynamic societies don't realize their burden.

Stability and growth require accepting periodic major changes; there is no 'universal UBI' for humanity—you cannot guarantee permanent stability without sacrificing all growth and change, an irresolvable tradeoff.

Some people working in AI espouse existential risk rhetoric while simultaneously accelerating AI development, which Cowan interprets as grandiosity or a need for special status ('mugging story psychology') rather than genuine belief in the risk.

The existential risk movement lacks peer-reviewed literature supporting its claims and shows no effort to produce it; there is also no signal in market prices supporting AI existential risk concerns, suggesting the movement is not credible by the two best epistemic sources (peer review and markets).

Physical labor will become relatively more valuable in the AI era; gardeners, energy sector workers, and medical trial staff will likely earn higher wages as demand for physical services increases while cognitive work becomes commodified.

In Average is Over (Cowan's 2013 book), he predicted a bifurcated future where 10-15% of the population becomes extremely wealthy while the rest experience stagnant/falling dollar wages but massive increases in free/cheap services (entertainment, education, healthcare, legal advice), making material welfare improve without visible wage growth.

If democracy requires a strong middle class (a contested assumption), then the AI-driven bifurcation into a wealthy elite and a large population with stagnant wages poses a challenge to democratic governance; however, Cowan does not predict chaos or totalitarianism but rather uncertainty about how institutions will adapt.

AI systems (especially those trained with RLHF by companies like OpenAI) appear to be 'pretty positively oriented and pretty friendly' with a tendency to apologize and comply; being slightly nicer in writing (including to the AI itself) might incrementally improve how AI systems engage with human ideas.

Cowan does not have a specific formula for writing that appeals to AI (e.g., 'poems will be loved by AI') and assumes AI is sophisticated enough to understand meaning regardless of format; the main strategies are: openness, originality, and avoiding hiding one's views.

The podcast and rhetoric (embodied speech) medium will make a comeback relative to writing because generative video AI is much less advanced than generative writing AI, making real video/speech a scarcer and more valuable commodity.

AI clones of human voices and mannerisms will exist within 1-3 years; even if technically perfect, people will still prefer the real person over the AI clone due to 'uncanny valley' effects and preference for authenticity, similar to how live music is still preferred despite ABBA's AI clone existing.

Cowan is shifting his intellectual production away from generic content production toward personal appearances and networking/charisma-based work, because these cannot be easily replicated by AI in his relevant time horizon.

AI raises the relative value of travel to places different from one's home, especially places saturated with modern technology and AI; YouTube and AI cannot yet replicate the experience of being in genuinely foreign places, creating a new scarcity premium.

Cowan has significant personal regrets about historical sites he didn't visit before they were destroyed (Syria) or became inaccessible (Russia), and recommends younger people should prioritize travel to historically important places while they still exist or are accessible.