What was actually worth it.

Mubien Ahsan

My own essay on whether independent assurance can actually make AI trustworthy to the institutions now betting on it. I argue that AI auditing sits roughly where financial control attestation stood in 2003, and that four unsolved problems decide whether it becomes a real discipline or a rubber stamp: what you measure against, what counts as evidence, who is qualified to judge, and whether the auditor is truly independent. Built on OSFI's model risk guideline, the EU AI Act and ISO 42001 rather than on opinion.

Michael A. Singer

Singer's argument is that the voice narrating your life is not you, it is just something you can notice and stop obeying. That one idea changed how I handle a busy head, which turns out to matter as much for building things as any technical skill. The least technical book on this page and probably the one I have used most.

Simon Sinek

The practical companion to Start With Why: less manifesto, more workbook, with an actual process for pulling your purpose out of your own stories rather than inventing a slogan. It gave me the words for what this site is for, which is why the About page reads the way it does. Useful if you can explain what you do but go vague on why.

Karen Hao

Hao reported inside OpenAI and the wider industry for years, and the result reads less like a technology book than an account of how power and money concentrate. The chapters on the data workers behind the models, and on what these systems consume, are the ones I have not been able to unsee. Uncomfortable if you are enthusiastic about AI, which is exactly why it earns a place here.

Ali Abdaal

Abdaal's claim is that feeling good is not the reward for being productive but the source of it, and he builds that into energizers, unblockers and sustainers instead of a lecture about discipline. I do not buy every study he cites, but it reframed evening projects as something to enjoy rather than grind through, and that reframe is the only reason this site exists.

Dario Amodei

Amodei's January 2026 follow up to Machines of Loving Grace. Where that essay made the case for the upside, this one walks through the risks: misuse, misalignment, authoritarian power and economic disruption. What makes it worth your time is that he refuses both doom and cheerleading, and then puts an actual plan on the table. I am not convinced by every prediction, especially the timelines, but it is the most serious writing on this I have read.