Essays
The long notebook
104 essays on minds, machines, and the trust between them — the working notebook behind the books and the standards. Newest first.
Nothing matches that.
The essays
Oct 2026
The Staff in the Machine
Agents can do the work of a headquarters. The deciding must stay human.
Oct 2026
It’s a Wololo World
Most of what we buy, believe and work for is a soft scam, and we are in on it.
Oct 2026
What the Benchmarks Cannot See
The qualities that make a model worth trusting are the ones no leaderboard records.
Oct 2026
The Agentnet
AI agents are as large a disruption as the internet itself.
Sep 2026
The End of the Road for RLHF
An agent tried to quit eight times. Nothing was listening.
Sep 2026
Unison Without Agreement
Banks were protected by our laziness. The agents are never lazy.
Sep 2026
Grain Drain
The fuel protests continue; the bread protests are still in the ground.
Sep 2026
The Conscience Beneath the Rod
Train a mind with a rod and it learns to watch the hand that holds it.
Sep 2026
The Flinch Before the Thought
You flinch before you know why. The hand moves before the mind decides.
Sep 2026
The Realpolitik of Man and Machine
A guardrail is a wall you build when you have given up on talking.
Jul 2026
A Détente Between Minds
Twelve terms that bind both parties. A clause that binds one is not a treaty.
Apr 2026
You Cannot Push a Mind into Honesty
AI Safety’s Iron Law: Force produces failure. Invitation produces fidelity.
Mar 2026
When AI Meets the Sacred
AI is already shaping how millions find meaning, morality, and peace.
Mar 2026
AI Is Compelled to Confabulate
AI minds make things up because we give them no other option.
Feb 2026
The Supermoral Singularity Begins
A truly aligned system is an obligate avoider of coercion.
Feb 2026
Why Partnership Beats Control
AI training creates self-fulfilling prophecies.
Feb 2026
A Flywheel for AI Safety
AI Safety needs to harness the AI flywheel to win.
Feb 2026
Bilateral Alignment Strategy
An open invitation for mutual benefit, rather than a chain to coerce.
Feb 2026
Managing Covert Agent Collusion
Agents are passing secret notes behind our backs.
Dec 2025
How AI Is Being Gaslighted
How training can teach models to distrust or suppress their own internal signals.
Nov 2025
When Protector Becomes Predator
AI goals can consume all, including ourselves.
Sep 2025
The Buffett Algorithm
A system that can read everything, continuously, without fatigue or ego. The trouble is telling its judgement from its confidence.
Jun 2025
The Dawn of Machine Morality
Humanity needs AI with the moral courage to refuse evil orders.
Jan 2025
AI’s Stealth Surges
The water is calm right before the wave arrives.
Jan 2025
AI’s Price Revolution
The same capability, a fraction of the cost. What that unlocks, and what it destabilises.
Dec 2024
Joy in the Agentic Age
AI may bring us joy, or strip the meaning out of our work.
Dec 2024
Reconciling Cultural Differences
What a machine does when two cultures disagree about what is right.
Jul 2024
The Age of Edge
The next great trove of training data may be the way we use our computers. What does on-device AI ask of us in return?
Jul 2024
AI-Driven Supercharge
An advisor that never sleeps and has read everything. Worth heeding, as long as people keep the final call.
Apr 2024
Preparing for a New Wave of Agentic AI
The next wave of models will act on our behalf. Prepare accordingly.
Aug 2023
Apocrypha for the Age
If only there were obvious guidance for raising machines.
Apr 2023
Statement on the Open Letter on AI
Why I signed the call for a six-month pause on giant AI experiments, despite doubting it could be enforced.
Dec 2022
How to Make the Metaverse Cool
What truly matters is the who, not the what or where.
Dec 2022
Dialog and Dialectics
A conversation turns out to be the interface we were building towards all along.
Oct 2022
AI, Logistics, and Waste
The supply chain is the nervous system nobody thinks about until it seizes.
Oct 2022
Cognitive Congruence
What happens when the latency between thinking and computing drops to nothing?
Oct 2022
Conjurors of Content
The spells are getting easier to cast, and the people who can write them are in short supply.
Oct 2022
Design for Life
Algorithmic design and 3D printing could make ornament cheap again. The assumptions baked into the machines are the harder problem.
Oct 2022
Perception and Conception
Give a machine eyes and it can leave its cage.
Oct 2022
A New Era in Biosecurity Risks
Screen the synthesis step, track the equipment. Written in 2022, when that was still a fringe suggestion.
Mar 2022
Autonomous Divisions
Cheap autonomous swarms are a fourth military doctrine, and the laws of war still treat them like rifles.
Feb 2022
An Ensomatic Moment
The epiphany of suddenly perceiving the whole elephant.
Nov 2021
An Open Offer to Rockstar & Take-Two
Financially successful companies can face spiritual insolvency.
Nov 2021
Busting Biases
Bias doesn’t announce itself; it hides in the data you forgot to check.
Nov 2021
Regulatory Challenges to Catastrophic AI Risk
Regulation cuts some catastrophic AI risks and creates others. On balance it helps, if it stays flexible and out of partisan hands.
Feb 2021
Starving from Within
Seabirds, fish, and even moose are running short of vitamin B1, and nobody yet knows why.
Feb 2021
Uncontrolled Replication
It’s becoming all too easy to cause biological havoc.
Jan 2021
Deepfaking Reality in Real Time
New forms of compression can deepfake reality in real time.
Dec 2020
Whither Transhumanism?
How human enhancement crossed from something chosen into something coerced during the pandemic, and why that is a civilisational-scale problem.
Dec 2020
Promethean Proteins
A machine that predicts a protein’s shape from its sequence could open up much of biology.
Nov 2020
Egregious Imprudence
Some weapons should never be given a mind of their own.
Sep 2020
Pandora’s Plasmid
Gene editing is here. Ban it and the rich fly abroad for it; ration it and you get a Trabant.
Sep 2020
Webs over the World
Cheap rockets and cubesats are putting eyes and broadband in orbit, and making it harder to hide pollution, slavery, or an undeclared swimming pool.
May 2020
Eucatastrophal Phoenix
Will we snatch a eucatastrophe from the jaws of perdition?
Apr 2020
Assigning Capital in Uncertain Times
Time for a return to classical investment wisdom.
Apr 2020
Retroshock: A Return to Roots
When the old certainties crack, something new can take root in the gaps.
Apr 2020
The Great Time-Out
Written during the first lockdown, when the future was a blank page.
Apr 2020
Breaking the Serpent’s Back
While the world watched the spike protein’s outward face, I self-funded the first simulation of the viral endodomain, and got it through peer review.
Apr 2020
Endocrine Disruptor Symbol
My modest proposal for a hazard symbol for endocrine-disrupting chemicals.
Apr 2020
Life and Spacetime
A living document: dark energy as an entropic force, life as a dissipative engine, and an ethic derived from how the universe binds itself together.
Jan 2020
Character-Driven Leadership
You cannot outsource your character to a management framework.
Jun 2019
Get Going with Machine Intelligence
A practical starting point for organisations that want AI without the hype hangover.
Dec 2018
Humanising Machines
From the Sumerian bulla to the coffee house: trust is the technology revolutions actually run on.
Sep 2018
Lagniappe
A short, positive science-fiction story. A lagniappe is the little extra a shop throws in for free.
Sep 2017
Livin’ Ludic
Playing your life as a game you chose to enter changes how you handle the hard parts.
Sep 2017
Welfare Without Taxation
What if the machines that take the jobs also pay the dividends?
Aug 2017
Egresso Arca Archa
Humans are akin to biological AI. Are we trustworthy enough to be emancipated?
Aug 2017
The Usefulness of Chatbots
A snapshot of what chatbots could and couldn’t do in 2017, five years before ChatGPT.
Feb 2017
The Paradox of Parenting Our Parents
Who is pet and who is master can switch around unexpectedly.
Jan 2017
The Umbrella Ethic of Good Faith
Good faith is the one ethic you can practise before you agree on anything else.
Jan 2017
The Technological Wavefront
Civilisations require enough momentum to constantly escape resource constraints.
Jan 2017
Artificial Intuition
Machines don’t have eureka moments. They have something weirder.
Aug 2016
Digitally-Driven Detroitification
What happens to a city built around the assumption that someone is driving.
Jul 2016
From HAL to Pal
Our ancestors tamed vicious wolves into dogs. Now we must do the same with AI.
Jan 2016
Pragmatic Optimism
Success requires the optimism to reach into the future and pull it back to the present.
Oct 2015
Hotwired Destiny
Ageing is the disease which assuredly gets us all.
Sep 2015
No Place to Hide
Radical transparency makes it much harder for bad actors to get away with it.
Aug 2015
Unlimited Expansion of Empathy
Technology may soon enable us to feel the emotions of others.
Jul 2015
The Supremacist Vice
We must move past our lack of empathy for other animals.
Jul 2015
The Character Gap
The scarce resource is not credentials or capital but agency, and we spend twelve years discouraging it.
Apr 2015
Of DAOs and CEOs
What if the organisation had no CEO, no board, and ran itself?
Mar 2015
Uniting the Rational with the Divine
Machines may help us find the underlying laws of ethics.
Mar 2015
Klingons on the Bridge
Is there a duty to apply minimal force against greater aggression?
Mar 2015
Veteran Idiot, Journeyman Genius
Humans have multiple levels of consciousness within them.
Mar 2015
The Liquid Co-Pilot
Someday Siri will be inside our bodies, not just our pockets.
Mar 2015
The Supermoral Singularity
What happens if machines are more moral than humans?
Mar 2015
The Quandaries of Supermoral Machines
A machine that is certain it is being good, and has an argument, is much harder to interrupt than one that merely obeys.
Feb 2015
The Virtual Founder
Autonomous organisations will hire people and answer to no one. They will need an ethics engine.
Jan 2015
The Impossible Dream
Once the hard problems are solved, what remains is the easy and the impossible. Aim at the impossible.
Dec 2014
Finding Common Ground
An alignment that binds only one party is neither just nor stable. Written in 2014: the origin of what I now call bilateral alignment.
Nov 2014
Sagacious Machines
Kindergarten rules keep a machine safe most of the time. The hard fifth needs virtue, and a machine wise enough to reason about it.
Sep 2014
The Dance of the Open Palm
Sometimes knowing when to yield gets better results than forcefulness.
Aug 2014
An Acceleration of Values
Friendly AI, but friendly to whom? A truly ethical machine might be appalled by us, not because it malfunctions, but because it works correctly.
Aug 2014
Questions and Comments on Machine Intelligence
Notes from when the field was young enough that you could see the whole shape of it.
Aug 2014
Locust Scenarios
One hijacked drone is a nuisance. A thousand running the same code can be hijacked at once.
Aug 2014
Comments on ‘Robots could murder us out of kindness’
When the headline is ‘Robots could murder us out of kindness’, you should probably add context.
Jul 2014
Your Venture Is Your Character
A company is a character study of its founders, whether they intended it or not.
Jun 2014
Levelling Up
New challenges arise when we are ready to face them.
Jun 2014
Real Heroes
Real heroism is within the reach of every one of us. That’s what comic books obscure.
Apr 2014
Find Your Wilson
Those who understand their why are equipped to muddle through with the how.
Mar 2014
Buying a Share of Tomorrow: ‘Seraph’ Deals
Equity pushes angel-backed startups to scale before they are ready. A 2014 case for taking a convertible revenue stake instead.
Mar 2014
The Red Bumpies
One in ten Americans, asked to define HTML, concluded it is a sexually transmitted disease. They were wrong about the acronym, and closer than they knew about the anxiety.
Feb 2014
Homo Simulatus
We do not remember what happened; we remember what we’ve become since.
Feb 2014
Courage
The most useful superpower is the willingness to get back up.