The ledger

The journal, read as data.

116 essays, 157,524 words, and 651 margin notes since April 2024. This page reads them back: what I keep writing about, what I've stopped writing about, the few calls I've actually bet on, and the positions I keep coming back to.

116essays
15themes
537links between essays
15calls on the record
24positions

How this page is made

Nothing here is counted by hand. Every chart, list and number is generated by reading the archive itself: all 116 essays, with their margin notes, themes and references.

The reading is done with AI. I use Claude to go through every essay and pull out the themes it touches, the films, books and people it reaches for, and every forward-looking claim. The margin notes come out of the same read, connecting each essay to the ones before and after it. Every quote is checked word for word against the published text.

The judgment is mine. Each claim was researched against what has actually happened since, with sources, and I ruled on every one. A call is something most people didn't believe when I wrote it. A reading is something already in the air. Anything that can't be proven right or wrong was taken out of the scoring, and the convictions underneath became the positions.

It updates with every essay. When a new essay is published, it goes through the same read. Its notes are added, older essays get notes pointing forward to it, its claims join the record, and the trends and positions are recalculated. The page you're reading reflects the archive as of Oct 2026.

Trends

Where the writing moved

Share of essays touching each theme, across a rolling window of the last 20 essays. The Epilogue Economy carried 65% of my first 40 essays and 16% of the last 25. Agents and friction came from almost nowhere, and data and insight is the newest climber, rising from 4% to 32% over my last 50 essays.

Show as a table

Theme	First 40 essays	Essays 67–91	Last 25 essays
Agents & systems design	5%	12%	28%
Friction & agency	5%	12%	32%
Aging & the Epilogue Economy	65%	12%	16%
Data & insight	10%	4%	32%

Every theme

Rising, steady, fading

Each line is one theme's share over the same rolling window. The figure is the change between the 25 essays before and the last 25.

Data & insight ▲ +28 pts Friction & agency ▲ +20 pts Agents & systems design ▲ +16 pts Prompting & tools practice ▲ +12 pts Presence & the analog ▲ +12 pts Aging & the Epilogue Economy ■ +4 pts Identity & reinvention ■ +4 pts AI adoption & belief ■ +4 pts Future of work & hiring ■ +0 pts Culture & society ■ −4 pts Learning & habits ■ −8 pts Taste & creative judgment ▼ −12 pts Brand & marketing strategy ▼ −16 pts Storytelling craft ▼ −20 pts AI filmmaking & video ▼ −28 pts

Density

When I wrote

Essays published each month. The busiest was Aug 2026, with 13 essays.

Show as a table

Month	Essays	Words
Apr 2024	1	653
May 2024	3	3,383
Jun 2024	1	955
Jul 2024	1	2,175
Aug 2024	1	994
Sep 2024	1	1,233
Nov 2024	1	985
Dec 2024	1	774
Jan 2025	1	311
Feb 2025	1	1,752
Apr 2025	12	8,686
May 2025	8	8,114
Jun 2025	12	11,333
Jul 2025	10	12,516
Aug 2025	2	2,222
Sep 2025	6	9,527
Oct 2025	2	3,448
Nov 2025	5	5,338
Dec 2025	2	3,137
Jan 2026	3	3,068
Feb 2026	4	6,071
Mar 2026	2	2,389
Apr 2026	4	4,995
May 2026	3	5,850
Jun 2026	4	7,942
Jul 2026	2	1,777
Aug 2026	13	26,232
Sep 2026	5	8,769
Oct 2026	5	12,895

Conversations

How the essays talk to each other

Every arc is a margin note connecting one essay to another, 537 in all. Grey arcs reach back to an earlier essay. Orange arcs point forward, to where an idea went next (334 of them). Taller arcs span longer stretches of time.

We Optimized the Wrong Thing cited 29 times · Oct 2026
Expected to Arrive Experienced cited 21 times · Aug 2026
We Don’t Have an Aging Workforce cited 18 times · Aug 2026
Do You Believe in This Thing or Not? cited 16 times · Oct 2026
What AI Is Teaching Us About Being Human cited 13 times · Jun 2026

Calls

The bets on the record

15 times I wrote something most people didn't believe yet: dated, specific, and possible to get wrong. These are the only forward-looking claims scored as foresight. 4 are already showing early. None has landed yet.

Apr 2024Oct 2026

Early 4
Open 11

Early
Many workers will not know what to do with the time AI frees up, because their professional habits are built around urgency and reactivity.

UC Berkeley Haas research (Feb 2026) found AI tends to intensify work rather than free time, with workers filling the gains with more tasks. That fits my mechanism of habits built on urgency, but it is not yet direct proof that workers are 'lost' with the free time.

Time Well Spent · Dec 2025
Early
Any agency AI solution that takes months to develop will be outdated and underfeatured on arrival.

In March 2026 Gartner predicted that about half of agencies' proprietary AI platforms will be obsolete by 2029, and warned CMOs about lock-in. That supports the direction, but 'any' solution being dead on arrival is overstated, and no platform has yet been shown to have failed outright.

The Agency AI Solution You’re Waiting For Is Already Outdated · Jun 2025
Early
Brands that ignore 50+ founders will miss real innovation, as older entrepreneurs increasingly reshape innovation culture.

The rise of older founders is real: new-business rates for ages 45-64 are now well above those of young adults, and first-time founder applications from 55-64s rose 35% from 2015 to 2024, per Kauffman data. There is no evidence yet that brands are losing out by ignoring them, or that innovation culture has changed.

Signals in the Epilogue Economy #01 · Apr 2025
Early
Leaders will encode their judgment into AI twins, making strategic judgment a new form of organizational IP and legacy.

Real cases have appeared: Customers Bank's CEO had an AI clone deliver earnings remarks, and Meta is building a Zuckerberg AI that staff can consult on strategy (Fortune, April 2026). These are isolated, mostly publicity-driven experiments, and The Register reports pushback from employees.

What Would I Think? Building a Cognitive Twin with AI · Apr 2025
Open
As fluency, agents and embedded AI mature, AI usage hours may decline and separating AI time from work time may become impossible.

I wasn’t wrong about measuring Return on Inference · Sep 2026

Show the other 10 calls

Open
An economy obsessed with youth may discover it has been undervaluing perspective just as information becomes abundant, though institutions may take a long time to correct the mispricing.

Deputy US Marshal Raylan Givens Was Right · Aug 2026
Open
Inference allocation may shape the post-modern corporation the way capital allocation shaped the modern one.

Show me the verbs an organization uses with AI, and I can tell you what it thinks intelligence is for · Aug 2026
Open
Work will continue drifting from a 10/80/10 to roughly a 40/20/40 split between defining, making and judging.

We Don’t Have an Aging Workforce · Aug 2026
Open
A new role that holds intelligence, imagination, in-experience tech and content harvest together will emerge in experiential work; almost nobody does it yet.

The Presence Economy Is the Next Great Brand Frontier · Jun 2026
Open
AI will become a verb in everyday language, like Google or Photoshop.

If AI Becomes a Verb, What Kind of Verb Will It Be? · Jan 2026
Open
The next decade of work will be shaped by people and teams who protect their attention, not by those with the best models or most automation.

Time Well Spent · Dec 2025
Open
AI will let the next generation accelerate the formation of judgment, so experience will no longer take thirty years to build if standards are taught.

Teach Them What Good Looks Like · Nov 2025
Open
Companies that offer AI upskilling to older professionals now will gain loyalty, leadership and a real-world AI test group, not just digital literacy.

The AI Ambition Gap · May 2025
Open
Older consumers will become the first market rather than the last, as they already are in Japan and South Korea.

The $96 Trillion Wake-Up Call · May 2025
Open
Organizations that prepare for the Gen X milestone within 12 months will gain competitive advantages in innovation, market understanding and knowledge retention.

The Epilogue Economy in 2025 · Dec 2024

Readings

Reading the room

50 times I read what was already in the air and built on it. Most people believed these when I wrote them, so they aren't scored as foresight. What they show is how well the room was read: 5 have landed, 12 are showing early, 1 went wrong, and the rest are still open.

51232

Landing 5
Early 12
Open 32
Wrong 1

Read every reading

Landing
Remote and fractional work will increasingly let older professionals contribute as consultants, advisors and project-based team members as traditional employment evolves.
The Epilogue Economy · May 2024
Landing
Generative AI models will get faster and better and integrate seamlessly, and creative institutions will adapt in ways not yet visible.
The Game We Think We’re Playing · Apr 2025
Landing
Agencies that cannot deliver AI-augmented results will fall behind because clients already expect it.
Let Go · Jul 2025
Landing
When every brand runs through the same AI tools and templates, output will converge into sameness and bravery will become the differentiator.
We Gave Everyone the Same Tools · Aug 2025
Landing
The same approve button will come to authorize consequential actions like deploying apps, buying media, changing prices, moving money and hiring.
Accept, Allow, Yes, Send · Sep 2026
Early
Consumers 65+ will be the fastest-growing spending group in advanced economies by 2030, and companies that engage them via older professionals will gain revenue and share.
The Epilogue Economy · Jul 2024
Early
Five forces, led by AI and demographic shifts, will impact every job within three years.
My (continued) advice to those over 50 · Aug 2024
Early
With fewer young workers and more older adults staying in, labor markets will flip toward phased careers, encore roles and age-diverse leadership.
The $96 Trillion Wake-Up Call · May 2025
Early
If the last decades made youth the cultural engine, the next few will make longevity the strategic one.
IMF Report - Not a Burden, a Growth Market · Jun 2025
Early
Simply returning to the office will not close the post-COVID leadership gap without structured development and mentoring.
The Leadership Gap No One Talks About · Jul 2025
Early
If 'good enough' becomes the standard, the industry will slowly and irreversibly lose the people able to do hard strategic and storytelling work.
When Did “Good Enough” Become Good Enough? · Jul 2025
Early
Investor attention (led by UBS) to longevity will pull industries toward building for older consumers.
The Epilogue Economy Wasn’t a Theory · Jul 2025
Early
The AI interface will shift beyond the text prompt to voice, image, annotation and possibly gesture, while the human questions stay the same.
Ghosts in My Notebooks, 0 to 1 thinking, and how I have moved past these · Sep 2025
Early
In the coming years the real workforce and market divide will be between GenAI thinkers and non-GenAI thinkers, not between age generations.
GenAI · Sep 2025
Early
Hiring will need to evaluate AI fluency and marketing will segment by mindset rather than age.
GenAI · Sep 2025
Early
Live, physical experience is the next brand channel, with experiential budgets rising as trust in online content collapses.
The Presence Economy Is the Next Great Brand Frontier · Jun 2026
Early
As AI makes production cheap, experienced people with contextual judgment may become more valuable because of the machines, not despite them.
We Don’t Have an Aging Workforce · Aug 2026
Open
Organizations that fail to harness older workers will be left behind amid unprecedented demographic change.
The Epilogue Economy · May 2024
Open
Precedents set in 2025 as Gen X turns 60 will shape attitudes toward aging and work for decades, and Millennials will reimagine their own career longevity in response.
The Epilogue Economy in 2025 · Dec 2024
Open
Synthetic audiences will not replace authentic voices yet, implying they eventually might.
Meet the Synths · Jun 2025
Open
Agencies that waited for AI to stabilize will trail those that moved while it was messy; builders, not dreamers, will lead.
Dreams Make Good Stories · Jun 2025
Open
Within a decade, the organizations that thrive will be those that built cultures of trust and used AI as a lever, not those that kept people on tight leashes.
AI Leadership · Sep 2025
Open
Organizations that adapt to AI will be those retraining everyone to think differently, not those hiring AI specialists.
To Question Is to Answer · Oct 2025
Open
The AI wave will follow the same pattern as previous ones: noise fades, and the advantage goes to quiet integrators rather than the loudest adopters.
I’ve Lived Through More Than One End of the World · Dec 2025
Open
Treating AI video as a cheap shortcut will end, replaced by treating it as a creative systems discipline that must be designed and governed.
The $29.95 Commercial · Feb 2026
Open
The gap between people who can generate footage and those who can structure visual experience will keep widening even as models improve.
AI Didn’t Kill Cinematography · Feb 2026
Open
As AI accelerates content volume, the competitive advantage will shift from making more content to making meaningful, structured stories.
How four decades in the creative industry — and five years experimenting with AI — led to the Story Engine behind RockPaperScissors · Mar 2026
Open
Brands that win the next era of storytelling will be those that experiment most frequently with small films and story prototypes, not those with the biggest campaigns.
The Bucket Hat Experiment · Mar 2026
Open
A growing share of people making AI films will never have experienced traditional film craft, reshaping how the medium is learned.
Notes from My Workshop at Spikes Asia · Apr 2026
Open
The hybrid shoot-less, build-in-AI production model will be what matters most to brands, agencies and independent creators over the next two years.
I stopped writing and went quiet to rethink how I work · Apr 2026
Open
Vertical micro-drama combined with hybrid production may become the model that makes brand storytelling at scale actually achievable.
I stopped writing and went quiet to rethink how I work · Apr 2026
Open
AI will disappear into infrastructure like electricity, and attention will shift to what it cannot replace: trust, connection and presence.
What AI Is Teaching Us About Being Human · Jun 2026
Open
AI and the Presence Economy will accelerate each other: as synthetic content grows, lived experience becomes more valuable.
What AI Is Teaching Us About Being Human · Jun 2026
Open
AI-driven first-pass recruitment screening will amplify historical age bias at scale unless audited.
Ageism Awareness Isn’t the Finish Line · Aug 2026
Open
Organizations that automate the work juniors learned through, while pushing out experienced staff, will eventually face a shortage of people with independent judgment, a trade AI output can hide only for a while.
Expected to Arrive Experienced · Aug 2026
Open
In AI-native workflows, the valuable asset will increasingly be the structured, machine-readable thinking upstream of final deliverables.
What the Hell Is an .md File, Anyway? · Aug 2026
Open
Access to AI will soon stop being an advantage; the race to acquire capability is nearly over and the next race is inference allocation.
Return on Inference (ROI) · Aug 2026
Open
As AI makes Option A (and contrarian Option B) cheap at industrial scale, the ability to reframe the question will become one of the era's most valuable creative and strategic skills.
In Defense of Option C · Aug 2026
Open
Because agents act rather than just talk, instruction design will become less like writing a creative brief and more like designing a behavioral system.
Taylor Swift Taught Me How To Write Better System and Agent Prompts · Aug 2026
Open
Machines will keep taking over the transactional hours of work whether organizations plan for it or not.
I Went for a Walk in Bucharest and a Meeting Broke Out · Sep 2026
Open
The decline in young adults' relationships is structural, not cyclical, and won't recover on its own.
An Algorithm Walks Into a Bar · Oct 2026
Open
The flat AI playing field, where everyone has the same models, will not last; believers will pull ahead and holdouts will usually lose.
Do You Believe in This Thing or Not? · Oct 2026
Open
Once every brand offers effortless functional experiences, convenience becomes table stakes and brands will have to compete on the meaning of having the thing.
We Optimized the Wrong Thing · Oct 2026
Open
In a world flooded with effortless content, asking something of the audience may become marketing's differentiator.
We Optimized the Wrong Thing · Oct 2026
Open
As effort is removed from ordinary life, chosen effort (handmade goods, rituals, participatory experiences) will grow more valuable.
We Optimized the Wrong Thing · Oct 2026
Open
Flying cars will not become mainstream personal transport, because regulation, infrastructure, energy and insurance stand in the way long before engineering does.
Spaceships, Flying Cars, and HAL · Oct 2026
Open
The flying Tesla Roadster will miss its 2028 production date.
Spaceships, Flying Cars, and HAL · Oct 2026
Open
As AI generates unlimited content and automation removes effort, human-made work, physical presence and chosen difficulty will command a premium.
Spaceships, Flying Cars, and HAL · Oct 2026
Open
As AI makes it cheap to analyze what content itself reveals, reading an organization's published archive for how its attention has moved will emerge as a practice distinct from performance measurement.
I Didn't Know What I Was Writing About · Oct 2026
Wrong
Widowhood will become more common as lifespans lengthen and women outlive men, reshaping consumption and community.
After the We · May 2025

Positions

What I keep arguing

24 convictions that run through the archive, each argued across several essays. The bar is how many. 13 of them have moved since I first wrote them, and the page for each says how.

Show the other 19 positions

Every position Where each one started, the essays that argue it, and where it moved.

Insights

What I noticed

39 observations about how something actually works, the kind you can take out of the essay and use somewhere else. Everything else that sounded like an insight turned out to be one of my positions, said well.

Production budgets were never just creative costs; they were a logistical negotiation with physics.
The $29.95 Commercial · Feb 2026
Convenience wins internal arguments because it is measurable, while anticipation and going out of your way have no dashboard.
We Optimized the Wrong Thing · Oct 2026
Systems that always answer demo better than honest ones, which bakes overconfidence into what gets bought.
What an Agent Does When It Doesn’t Know What To Do · Sep 2026
Most talent teams can break down hiring funnels by gender within an hour but not by age, which shows which biases they chose to watch.
Ageism Awareness Isn’t the Finish Line · Aug 2026
Asking what competitors are not providing is a sharper strategic question than asking what customers want.
In my 1-1 sessions, I ask one question, and I am always surprised by the answer · Nov 2024

Show the other 34 insights

Rituals don't shield us from change; they give us a rhythm so we notice when something shifts.
“We Just Decided To” · Apr 2025
Experienced staff were often cut not because they couldn't adapt but because their memory of past failures made new initiatives harder to sell internally.
Artificial Intelligence Gave Us Answers · Jun 2025
Experienced audiences are not confused but selective; they want recognition, not instruction.
Explainer Mode Is Lazy Marketing · Jun 2025
Most portfolios reflect the safe briefs and conservative clients a creative was given, not the creative's actual thinking.
We Gave Everyone the Same Tools · Aug 2025
Agency systems were built to ration value, so they cannot cope with AI's abundance.
Beautiful Tools, Broken Systems · Oct 2025
Before prompting, decide whether you want proof, possibility, perspective or provocation, because each needs a different question.
To Question Is to Answer · Oct 2025
Organizations misread temperament as dysfunction and ask Settlers to act like Explorers as if temperament were a setting.
The Explorer, the Settler, and the Future of Work in the Age of AI · Nov 2025
Each tech wave made something faster or more abundant, not better, and punished those who outsourced judgment to it.
I’ve Lived Through More Than One End of the World · Dec 2025
Once likes arrive, the question shifts from whether the work is good to whether it landed, and judgment becomes crowdsourced.
Bad Work Is Fine · Jan 2026
AI did not invent slop; it exposed how much performative work organizations had already normalized.
On AI · Jan 2026
Photoshop's arc shows a tool maturing when restraint becomes the skill and choosing not to intervene matters as much as intervening.
If AI Becomes a Verb, What Kind of Verb Will It Be? · Jan 2026
People's unease about AI comes from its proximity to the faculties tied to meaning, not its power.
Do Not Underestimate Your Place in History Right Now · Feb 2026
AI-native creatives arrived after the industry playbook was already burning, so they carry no inherited sense of what is impossible.
Oh, The Places We Will Go · Apr 2026
A soundtrack tells an audience how to feel but rarely tells them where they are; place comes from ambient sound.
My recent AI video moved like a shark, but sounded like a goldfish · May 2026
Advertising is excellent at optimizing existing ideas and has almost no infrastructure for tolerating the long period when someone is finding their voice.
Thom Yorke Just Said the Quiet Part Out Loud · May 2026
What matters about young people's AI use is less what they use it for than what they use it instead of.
What AI Is Teaching Us About Being Human · Jun 2026
For the first time you can think out loud to something that thinks back, with no social consequence.
What AI Is Teaching Us About Being Human · Jun 2026
Companies automate the middle of work while offering packages to the people best at the bookends, then wonder why output feels off.
We Don’t Have an Aging Workforce · Aug 2026
Opening PowerPoint chooses the shape of the thinking before any thinking is done.
What the Hell Is an .md File, Anyway? · Aug 2026
Phone books were public but frictional; the privacy shift is less about exposure than about aggregation and retrieval.
I am who I remember myself to be · Aug 2026
Best practice is a vast library of Option As, and contrarian Option B is still controlled by the frame it rebels against.
In Defense of Option C · Aug 2026
A bookstore shelf is itself a recommendation system; what distinguishes it is that it offers a possibility without a conclusion.
In Search of Maybe · Aug 2026
Specific details in a lyric work as retrieval, loading a whole system of meaning in two words; prompt details should do the same work.
Taylor Swift Taught Me How To Write Better System and Agent Prompts · Aug 2026
Decorative specificity makes a prompt sound like it knows what it wants; consequential specificity changes what the model does.
Taylor Swift Taught Me How To Write Better System and Agent Prompts · Aug 2026
Walking side by side removes the 'head of the table', so you can't present to someone, only talk to them.
I Went for a Walk in Bucharest and a Meeting Broke Out · Sep 2026
Institutions need spouses while people are looking for partners, a vocabulary gap behind much of the grumbling about state matchmaking.
An Algorithm Walks Into a Bar · Oct 2026
Every forecast is built from the forecaster's present, so the details that date it fastest are the things that felt most modern when it was made.
Spaceships, Flying Cars, and HAL · Oct 2026
Futurists overestimate how fast we will change the physical world and underestimate how fast we will change the informational one, because physical change runs into economic and social resistance that information mostly slips past.
Spaceships, Flying Cars, and HAL · Oct 2026
Prediction tends to get the device right and the use wrong: fiction had the communicator and the tablet early but missed how much of life would converge on them.
Spaceships, Flying Cars, and HAL · Oct 2026
Technology forecasts track what becomes possible, while adoption tracks what people want, and the two diverge often enough that feasibility is a poor predictor of the future.
Spaceships, Flying Cars, and HAL · Oct 2026
People notice when they change their minds but rarely when they change what they pay attention to, which is the slower and more consequential shift.
I Didn't Know What I Was Writing About · Oct 2026
An idea can survive by changing its label, so tracking a named theme can show it fading while the underlying argument carries on under other words.
I Didn't Know What I Was Writing About · Oct 2026
Content can be asked three different questions: how a piece performed, what an organization has been talking about, and how its attention has changed over time. Only the first is routinely measured.
I Didn't Know What I Was Writing About · Oct 2026
A published archive is not a window into an organization's thinking, but it is a reliable record of what it repeatedly chose to make visible.
I Didn't Know What I Was Writing About · Oct 2026

Lines

How I said it

The sentences that stand on their own. Each one opens the essay it came from.

Show the other 20 lines

Range

What I reach for

274 distinct films, books, songs, people, places and moments in history used as a way into an argument.

People92 Films36 History35 Places31 Music23 Books20 Science & research16 Television12 Sport9

Expand the range Every name, and the essays it appears in.

Reading the data

What it suggests

The center of gravity moved. The archive began as an Epilogue Economy project and became a working notebook on how AI changes the work. Data & insight, Friction & agency, and Agents & systems design gained the most ground in the last 25 essays.
Some threads went quiet. Taste & creative judgment, Brand & marketing strategy, Storytelling craft, and AI filmmaking & video have faded in the most recent essays. Quiet isn't finished. These are the obvious places to return with fresh evidence.
Mostly a reader, sometimes a bettor. 50 readings of the room against 15 real calls. The readings show judgment. The calls are where foresight gets proven, so the next ones should be fewer, sharper and dated.
Rising themes with no bets. Prompting & tools practice and Presence & the analog are where the writing is moving, yet almost no calls sit there. That's the place to go on the record next.
The hubs hold it together. We Optimized the Wrong Thing and Expected to Arrive Experienced are the essays the rest of the archive points to most. New readers should start there.