Essays
Longer thinking about AI and the people who have to live with it. History gets pulled in whenever it helps.
Read essaysFrom a desk in Chicago
Dinand
Tinholt
AI Consultant | Researcher | Writer | Builder
I spend my days helping companies put AI to work. Most evenings end in a history book, and this site is where the two keep bumping into each other.
01Writing
Reading Notes
What I took away from the books and papers that changed my mind about something.
Browse notes02Research & Experiments
Research
Simulations I run when I want to know if an idea about AI survives contact with real numbers.
Explore researchExperiments
Apps and research systems I built. Two of them run right here in your browser.
Explore experimentsSide Projects
Things I build for the fun of it, like a downtown Chicago you can walk through at any hour.
See side projects03Background
About
Born in the Netherlands, at home in Chicago, with a lifelong habit of asking one more question.
Get to know mePlate I
A cabinet of questions
The collection
Everything I think about,
in one place.
My work used to scatter. Essays landed on LinkedIn and Medium while the simulations stayed on my laptop. Here they finally sit side by side, and you can see they keep chasing the same handful of questions.
It changes as I learn. Expect drafts.
- 115
- Published articles in the archive
- 07
- Apps and systems I built
- 07
- Sections to explore
Choose an orientation
What brought you here?
Pick the description that sounds most like you. Each route hands you the three pieces I would give you first.
Or surprise meRethink how the company works
You run something, and you want to know where agents fit before they start running it for you.
Move from idea to instrument
You build things, and you like your assumptions out in the open where someone can poke at them.
Interrogate what the numbers mean
You want to see how a simulation was set up before you believe a word it says.
Take the long way around
You came for the detours, the spots where an old map or a forgotten novel has something to say about software.
I swap these out every so often. Each one is a good door into the rest of the archive.
AI oversight
Skin in the Game Made the AI Supervisor Worse
A controlled experiment on the cost of wrongful blocks and the incentives that shape an AI reviewer.
What changes when caution itself becomes costly?
Enter the inquiryPlate 02 · Simulated markets
Building a Fake Economy Run by AI Agents to Negotiate Trade Promotions: Lessons from a CPG Simulation
Manufacturer and retailer agents negotiate promotions inside a simulated market with competing incentives.
Read the piecePlate 03 · Maps & models
The past as a compass
Old maps and new technologies invite us to examine the assumptions behind what we see.
Read the piecePlate 04 · Agent memory
How to Make AI Agents Accurate: Stop Treating Memory Like Chat History
A practical approach to keeping an agent’s constraints and current decisions clear during long sessions.
Read the pieceFrom PublicationsSeptember 2021–September 2026
Five years of ideas
in motion.
One piece from each year, so you can watch the thinking move. Next to it sit a few talks I gave, each checked against the organizer’s own page.
Year by year
One selected work per year2026Accountability
Skin in the Game Made the AI Supervisor Worse
A controlled experiment on the cost of wrongful blocks and the incentives that shape an AI reviewer.
Read the overview2025Context
How to Make AI Agents Accurate: Stop Treating Memory Like Chat History
A practical approach to keeping an agent’s constraints and current decisions clear during long sessions.
Read the overview2024Complexity
From Chaos to Managed Complexity: Adopting an Ecosystem Mindset in Data & Analytics
A case for treating modern data environments as evolving ecosystems rather than trying to force every source and decision into one rigid center.
Read the overview2023Liminality
The Liminality of Data: A Glimpse Between the Thresholds
What the idea of liminality can teach data and AI leaders about working between an established order and one that has not yet taken shape.
Read the overview2022Culture
Creating a Data-Powered Culture
A durable starting point for making data part of everyday decisions: connect technology, shared habits and organizational confidence.
Read the overviewI checked each appearance against the organizer’s public page on September 21, 2026.
Across every sectionOpen questions
The threads I keep
returning to.
Four questions keep turning up, whatever I happen to be writing or building at the time.
01Judgment
Where should human judgment stay visible once systems begin to act?
Start with what happened when an AI reviewer had to answer for its past mistakes.
Follow the question02Memory
How much of yesterday should an agent carry into today’s work?
Long sessions drift. This is how I keep an agent anchored.
Follow the question03Resilience
How do we rehearse for disruptions we cannot predict?
I threw thousands of random shocks at an inventory planner to find out.
Follow the question04Representation
What makes a confident image or answer feel true?
John Berger asked it about paintings in 1972. It fits chatbots uncomfortably well.
Follow the questionFrom Reading NotesAt the reading desk
Notes I scribbled
in the margins.
These stay unfinished on purpose, because I keep going back to add to them.
The Society of the Spectacle
Guy Debord, representation, and a question about synthetic media.
Open the noteFrom PublicationsLatest published
Ideas out in the world.
AI oversight
Skin in the Game Made the AI Supervisor Worse
A controlled experiment on the cost of wrongful blocks and the incentives that shape an AI reviewer.
Explore articleAgents in supply chains
Building a Fake Economy Run by AI Agents to Negotiate Trade Promotions: Lessons from a CPG Simulation
Manufacturer and retailer agents negotiate promotions inside a simulated market with competing incentives.
Explore articleAgent engineering
How to Make AI Agents Accurate: Stop Treating Memory Like Chat History
A practical approach to keeping an agent’s constraints and current decisions clear during long sessions.
Explore articleHuman–AI collaboration
Symbiotic Intelligence Theory (SIT): Rethinking Human Potential in the Age of AI
A framework for thinking about human agency and the ways AI can extend our capacity to think together.
Explore articleFrom PublicationsThe ongoing conversation
From Medium
Selected articles. Checking for the latest posts…
Supply-chain resilience
Inventory Optimization Under Uncertainty: Broad random shocks trained a tougher inventory planner than a clever AI adversary did
An inventory-planning experiment explores which kinds of stress prepare a policy for unfamiliar disruptions.
Read on MediumAgent engineering
How to Make AI Agents Accurate: Stop Treating Memory Like Chat History
A practical approach to keeping an agent’s constraints and current decisions clear during long sessions.
Read on MediumAdaptive systems
The Next Frontier of Agentic AI: Context That Learns
How feedback from an agent’s work can inform the context it uses on the next task.
Read on MediumWhile this page is open it looks for new posts every five minutes, as long as Medium’s feed is up. Subscribe via RSS ↗
Elsewhere on the internet
Follow the work
as it develops.
Most weeks something new goes up on one of these.
41.8781° N
87.6298° W
Before you go
Start with whatever
makes you curious.
Pick whatever caught your eye on the way down and open it. If nothing did, the surprise button will choose for you.