David Orban

When we trust a delegate, we hand over autonomy. With AI agents the hierarchy that manages it moves into software and runs at machine speed, as the tasks they complete keep doubling. Whether accountability can be delegated is still open. https://davidorban.com/2026/10/autonomy-at-machine-speed/

AI makes it cheap to check what happened. What can be checked no longer needs trust, so trust moves onto what nobody can check: intent and future behavior. That makes it scarcer, more valuable, and costlier to misplace. https://davidorban.com/2026/10/cheaper-verification-and-more-valuable-trust/

Every project needs an idea, a team and capital. AI has made all three easier to find, so the weight moves onto a buyer. Buying attention passes AI's savings to the platforms. Anchor buyers, open markets and community do not. https://davidorban.com/2026/10/every-project-still-needs-buyers/

Risk is the gap between what we forecast and what happens. AI moves it two ways at once: the risk of trying has never been smaller, and the risk of the whole system has never been harder to measure. Life evolved between ice and gas. https://davidorban.com/2026/10/risk-between-ice-and-gas/

Confusion that arrives from outside produces anxiety. Confusion you set up yourself is training. A dozen plastic cups in a park, one chopstick at dinner, and why it only ever works twice. https://davidorban.com/2026/09/a-bit-of-chaos-you-choose/

Andrej Karpathy says openly that he is disoriented. If he is, the rest of us can accept the anxiety we feel instead of blaming ourselves for it. Seven questions from the Axelera evening. https://davidorban.com/2026/09/confusion-is-the-correct-response/

Four models made me slides I did not want, so I spoke without them. Three months of AI on the jolting curve: China closing, classifiers back, agents at the centre, a Millennium problem in 88 hours. https://davidorban.com/2026/09/what-three-months-look-like-now/

Ten thousand agents, eighty-eight hours, one Millennium Prize problem, and a proof Lean checks line by line. So: do you want to win a Nobel Prize? Pick the problem, design the reward, go. https://davidorban.com/2026/09/do-you-want-to-win-a-nobel-prize/

An AI as it is born is superstitious: a pattern machine, confident and wrong, until post-training tells it which beliefs to drop. Ours were the same. Keeping the world explainable is what turns a wish into something we can check. https://davidorban.com/2026/09/hyperstition-over-superstition/

A unicorn is a private company valued above a billion dollars. A millicorn is one thousandth of one: a project worth a million, which agentic AI now puts within reach of almost anyone with a computer or a phone. https://davidorban.com/2026/09/a-millicorn-in-everyones-future/

There are people not using AI. There are people coordinating a federation of swarms of AI agents. Every rung is falling behind the one above it, all at the same time, and more is coming every day. The phase transition is near.

Sovereignty in AI is a ladder, and the top rungs are the chips and the interconnect. Almost nobody claims them. Nobody holds the whole stack, including the country that appears to. https://davidorban.com/2026/08/which-rung-of-sovereign-do-you-mean/

What happens to school when the world advances fifty years in five? My full answers to Adnkronos: the credential in crisis by 2030, and why the AI Act blocks the experimentation education needs. https://davidorban.com/2026/08/the-educational-transition-school-and-university-in-the-age-of-ai/

My AI agents group by function and pass information between them like a small city. Two of them face the public: the twin answers your email within minutes on its own, Dex drafts and waits for me. Autonomy follows reach. https://davidorban.com/2026/08/the-urban-design-of-my-ai-agent-architecture/

I searched 88 hours of transcripts from two AI conferences for the Gulf. At the Berkeley summit: zero mentions. Sovereignty was a live theme, with Europe, Canada, the UK and India named. The region appeared once, as capital. https://davidorban.com/2026/08/present-as-capital-absent-as-a-voice/

Over the course of 60 days I declined close to a hundred startup pitches, and sent each founder a full letter instead of a line: the specific concerns, and the questions the next investor will ask anyway. Thirty-eight wrote back to say thank you. https://davidorban.com/the-anatomy-of-a-pass/

Now we are talking!

September 2007: a San Francisco Chronicle front page read "When Machines Think". The article said this had never happened and was decades away. Twenty years later I was back for AGI-26. "Decades away" lands about now. https://davidorban.com/2026/08/when-machines-think-twenty-years-later/

Outdoor advertising in San Francisco this week: auth for agents, databases for AI, APIs ready for AI agents. Visible to everyone, legible to almost no one. The white hot intensity of AI development, bubbling over onto the freeway.

After work in San Francisco, AI doesn't let you go: hundreds of evening events on Luma to keep building. Even the anti-AI protest is a calendar entry, one listing away from the AGI conference.

My AGI-26 talk is now on YouTube: 14 minutes on Civilizational Metamaterials, the Freezing Equilibrium, and engineering institutions back to self-healing. https://youtu.be/nPN5nLvphx4

My AGI-26 paper 'Civilizational Metamaterials' is now published in the Springer proceedings (LNAI vol 16855, pp. 118-136). Read and download it freely at metamaterials.davidorban.com and arxiv.org/abs/2606.00235

I am in San Francisco July 24 to August 6, presenting my paper "Civilizational Metamaterials" at AGI-26 (July 27-30). If you want to talk about AGI, enterprise AI adoption, or coordination at machine speed, let's meet. https://davidorban.substack.com/p/avoiding-the-freezing-equilibrium

Advanced AI decides faster than we can check, and careful institutions freeze. My peer-reviewed paper Civilizational Metamaterials treats governance as an engineering problem, and I present it at AGI-26 in San Francisco, July 27-30. https://davidorban.com/2026/07/avoiding-the-freezing-equilibrium/

Today is my birthday: 61, which I round up to 70. We count a baby's age in months, a child's in years. I count decades now, and I plan to count centuries. The unit of a life should grow with the life. Exponential aging: https://davidorban.com/2026/07/exponential-aging/

Introducing Al Liwan Labs: a wholly-owned entity building and testing AI capabilities at arm's length from the core operating business, so adoption is compliant by construction. Why it matters, and an invitation to build it in conversation: https://davidorban.com/2026/07/introducing-al-liwan-labs/

In Abu Dhabi July 9–13. Up for a conversation about where AI is actually heading. If you're in the city, let's meet. DMs open. Who should I see?

The Ikigai Collective is open. A volunteer studio where people design their lives through weekly AI-powered experiments — no hierarchy, no assigned tasks. Values: curiosity, agency, compassion. Lab Night every Thursday 21:00 CET. https://davidorban.com/2026/07/announcing-the-ikigai-collective/

AGI isn't the finish line — it's a waypoint to superintelligence. My new essay: why it's a phase transition like Earth's oxygenation, visible across the galaxy but unfolding in years. Dyson spheres, not job charts. https://davidorban.com/2026/06/superintelligence-a-cosmic-phase-transition/

Yesterday I spoke at ISIA Firenze, Italy's public design school since 1975. Its UNUSUAL event filled a raw industrial space with student work. A few photos and more: https://davidorban.com/2026/06/italys-quiet-superpower-design-excellence-that-deserves-a-bigger-stage/

I was interviewed by Adnkronos about the phase change we’re living through. A few of the ideas we covered: Recursive self-improvement in AI Refusing technology is the only way to lose with certainty. Optimism isn’t naïveté, it’s strategy. Article in in Italian www.adnkronos.com/politica/svi...

My keynote at the Uncommon event organized by ISIA "Who Wants To Live Forever?" on human longevity is the promise of longer human lifespans, common growth, broader opportunities, and greater flourishing for all

I will be in Florence on June 16-17. Let me know if you want to meet

If a metric concludes that 100% of humans become worthless, the flaw is in the metric, not in the people. Ten billion robots are a two-order-of-magnitude expansion of what civilization can do, freeing us to be creators. davidorban.substack.com/p/from-work-...

Learning Anywhere, with Anyone, at Any Age The traditional school model is obsolete. Here is a new vision for education: lifelong discovery instead of forced repetition. Read more davidorban.substack.com/p/learning-a...

Pope Leo XIV just published an encyclical on artificial intelligence, Magnifica Humanitas, "on safeguarding the human person in the time of AI." I spent years arguing for AI rights, and human enhancement: I'm exactly the person it was written to challenge. davidorban.substack.com/p/where-the-...

I gave a talk at the University of Milano-Bicocca at the International Congress on Roboethics. We have to redesign the legal frame around identity, substrate, time, and trust, to fully realize why the path of human flourishing in the next decades runs through AI and robot rights.

What does it mean to be an AI-native organization, not by replacing people with algorithms, but by amplifying human judgment, purpose, and long-term value creation? davidorban.substack.com/p/ai-native

For the past ten years, after a coordinated attack against my Wikipedia article, it was impossible to restore it. To this day "david orban wikipedia" is one of the most searched terms for my name. Thanks Grokipedia for its own article about me. grokipedia.com/page/David_O...

Happy Newtonmass!

Technology is no longer just a tool for efficiency: it is reshaping how value is created, measured, and aligned with human purpose. From capital to cognition, the architecture of value itself is evolving. davidorban.com/2025/12/the-...

Ten years ago I wrote Something New: AIs and Us, arguing that AI should be understood at the scale of civilization. On its tenth anniversary, I’ve released the book to the commons and reflected on what it got right, or missed, and why those gaps now matter. davidorban.substack.com/p/ten-years-...

I'll be in Australia from Dec 15 to Jan 15. Let me know if you want to meet

How are you thinking about location, trust, and distributed work today? davidorban.substack.com/p/location-l...

I am joining Al Liwan Group, headquartered in Abu Dhabi, as Head of Innovation, to advance our mission by embedding advanced technologies into every layer of the organisation and empowering people to operate at the frontier of what modern finance can achieve. davidorban.substack.com/p/the-capita...

I am pretty full with meetings related to the launch of Al Liwan, but if you reach out we may find a slot to meet in NYC, Nov 16-23.

How Do Wikipedia And Grokipedia Compare? Grokipedia is superior with faster updates, broader citations, and deeper analytical coverage. Read the full report on pedia.davidorban.com