History repeating itself, beat for beat. Feigenbaum's 1988 "The Rise of the Expert Company" reads exactly like today's AI economics papers: success stories, adoption barriers, job displacement concerns, calls for government infrastructure.

My latest post, "What Happens in the Gap". The one in which I show what actually has to happen between 'this technology works' and 'productivity explodes.'

The AI-Powered Web Is Eating Itself buff.ly/d1sXhtv The commons was already enclosed. We're just watching different landlords fight over who gets to charge rent. None of them actually want to tear down the fences.

You start in AI trying to build intelligent systems, realize you need to understand what intelligence is, and end up reading Damerow on how cuneiform tablet accounting practices generated the concept of number.

Current mood: Evil In Oslo by Hedvig Mollestad Trio music.apple.com/au/album/evi...

My latest Substack post, 'LLMs Are Following the Expert Systems Playbook—But the Score Is Different'. The one in which I explain why AI-skilled workers earn 56% more while 95% of enterprise AI pilots fail—and why both signal the same economic retreat.

It is fascinating how Damerow’s work strips away the "inevitability" of math. We often treat numbers like universal truths we simply discovered, but he shows they were hard-won cultural artifacts—tools forged to solve specific administrative and accounting crises in Mesopotamia.

Current mood: loveless by my bloody valentine on music.apple.com/au/album/lov...

Current mood: Houdini by Melvins music.apple.com/au/album/hou...

My new Substack post, the one where I trace the 30-year journey from 'motors as fire hazards' to 'networks as infrastructure." If you like thermodynamic efficiency, historical insurance data, and debunking standard economic myths, this one is for you.

An expert isn't someone who knows how things work; an expert is someone who possesses a comprehensive library of how things fail.

“aviation emissions could be halved without cutting journeys” buff.ly/rgbOMNm The study's real value isn't a roadmap—it's a metric. As Gössling suggests, requiring public efficiency ratings (like "F-grade airlines") might pressure change.

A characteristically thorough and grounded piece from @rodneyabrooks.bsky.social with his Predictions Scorecard, 2026 January 01. What makes it solid is his methodical tracking of predictions against reality, combined with his deep historical perspective (50 years in AI/robotics).

Today’s “LLM agents” are just prompt-chaining with tools—they simulate intention via recursive function calls, but have no real commitment, no persistent goal stack.

LLM seem to be following the expert systems trajectory. Wage premiums are the tell: we're paying domain experts to wield pattern matchers, not toolmakers to wield magic. Different technology, same gap between simulation and capability.

My latest, “2025: The Year that Wasn’t”, the one where I explain why 89% of organizations haven’t scaled AI agents: the problem isn’t infrastructure or training—it’s that simulated goals aren’t stable enough to justify the costs. A squirrel can’t be talked out of wanting birdseed. An AI agent can.

The Unbearable Joy of Sitting Alone in A Café | Candost's Blog buff.ly/c0E2rdi

In England, the best predictor of where the Industrial Revolution would take off wasn't the number of libraries, but the density of watchmakers and toolmakers.

Expertise is knowledge of failure modes, not operational principles.

Robodebt, UK Post Horizon, and now US CMS ift.tt/DMVtWXu

Current mood: Houdini by Melvins music.apple.com/au/album/hou...

My latest Substack post, "We Saw AGI on Mars", the one in which I explain why LLMs are sophisticated combinational engines, not AGI—and why the difference matters. Hint: it's about existential stakes vs. statistical surprises. The canals are in the mirror.

Current mood: Bees In The Bonnet by Hedvig Mollestad Trio music.apple.com/au/album/bee...

My latest Substack post, "The Present is Legible: Why your 2027 forecast is making you blind". The one where I show why squirrels, Wayne Gretzky, and Netflix all succeeded the same way: reading resistance instead of predicting futures.

New Kindle Feature Uses AI to Answer Questions About Books—And Authors Can't Opt Out - Reactor buff.ly/1X1sgj4 Can someone explain why why answering “Who is Lady Macbeth?” or “What happened in Chapter 3?” would require a license?

The grifters were prepare for the gov’s age verification, of the mark on the first day with fishing emails,

Current mood: Blues & Roots by Charles Mingus music.apple.com/au/album/blu...

(p.6, beside crypto/metaverse/LLMs) "Tech's grief cycle: Bitcoin (denial), Metaverse (bargaining), AI (acceptance + desperate hope)."

Shipping Containers: Malcom McLean’s Complicated Gift To Free Trade buff.ly/QM0zZHJ A creative angle on Malcolm McLean’s invention of containerized shipping, framing it through the lens of modern software development metaphors like Docker and containerization . Fun read!

Some of the (what looks like AI) animations of album covers in Apple Music is just a bit creepy. (Looking at you, The Pleasure Principle.)

The AGI panic is distracting us from present harms—Robodebt, UK Post Office Horizon, the algorithmic surveillance of warehouse workers—to focus on a science fiction future.

My latest Substack post, "The Three Grammars". The one in which I show the same pattern repeating: 1897 electrical code, 1925 auto loans, 1999 web standards. Each took 10-15 years. Now we need all three to converge simultaneously for heat pumps, EVs, and solar. That's the crisis.

Current mood: Heartattack and Vine by Tom Waits music.apple.com/au/album/hea...

What’s arguably most impressive about Zorn: he’s built a system for sustained creative output. The label, The Stone, the collaborative network—this removes friction that slows most artists down. Less ‘lone genius,’ more ‘genius at orchestrating creative ecosystems.’

New Knives (Wake Up Dead Man) out was fun. Worth going to the cinema for.

Current mood: Bone Machine by Tom Waits music.apple.com/au/album/bon...

We live in a state of epistemic anarchy. A return to elite gatekeeping won't work. Is the only hope to persuade the misinformed? www.conspicuouscognition.com/p/lets-not-b...

"only 20% of [customer service and support leaders ] leaders have reduced agent staffing due to AI" That's a bit of a change from the "mass layoffs due to automation" narrative and the "AI job apocalypse Gartner: buff.ly/1gz2X91

“I never practice. From time to time, I just open my case and throw in a piece of raw meat.” The late guitar ace who spurned the jazz world, climbed the pop charts and lived in terror of taking a solo | GuitarPlayer buff.ly/0iCv6Pt

(p.2, beside “Are you using AI to write this?”) “Turing test or tone-police?”

Steve Cropper, legendary guitarist for Booker T & the MGs, dies aged 84 | Soul | The Guardian www.theguardian.com/music/2025/d...

If you’re interested in why some trades resist full automation, the B1M’s piece on bricklaying robots is a perfect case study. Mortar is a non-Newtonian, stateful material; engineered adhesive is what makes robotics possible. A neat example of how we change materials to fit machines, not vice versa.

Latest: "The Agent That Wasn't There"—explaining why agentic AI security is a category error. We're building stateful pattern matchers (LLMs) and securing them as goal-driven agents. That architectural gap? That's the vulnerability. Think platypus, but for AI security.

Generative AI had the fastest “try it once” adoption ever because it’s free & one click away, but three years in, it has delivered no big productivity surge, no widespread job displacement, no new life-changing product, and no deep societal shift—just a very good tool that got hyped as a revolution.

Systems design 3: LLMs and the semantic revolution - apenwarr buff.ly/ep6TBtL Lots of fun ;)


(in the gutter, sideways)
 “Numan’s ‘Cars’ = first time the uncanny valley got a radio edit.”

A mathematical ceiling limits generative AI to amateur-level creativity buff.ly/d79dny9 Solid and interesting take by someone with a long track record of work in creativity

My latest on Substack, 'The Platypus in the Server Room', the one in which I compare AI researchers to Victorian naturalists inspecting a platypus pelt for stitches, and somehow this analogy holds up for 1,200 words.

Current mood: Hyper-Dimensional Expansion Beam by The Comet Is Coming music.apple.com/au/album/hyp...

Do I have it right that AI 2027 is now 2030ish, the hard takeoff is more a gentle loft, and the sharp left turn is now a gentle glide to the left?