Most people working on AI came to it from the technology. I came to it from people.
The technology is the easy half. Getting people to actually work differently is the hard half, and that is the half I do. At reMarkable the engineering team took what I built and made it the foundation of the company's AI work, across seven departments and three platforms. I do that for other people now, from one afternoon to a whole system.
- Huddly
- Jabra
- reMarkable
- Upwave
Why it answers well
and nothing changes
You have the same AI as everyone else. It answers well, and then you spend the afternoon correcting what it got wrong about your own company, because nothing ever told it. That is not a tools problem.
I start from one piece of work you already do that costs you something. Then the question most AI programmes never answer: what will you measure to say this worked? We take a baseline, change the work, and read the number afterwards. If it did not move, it did not work, and I would rather we both knew.
What I do
Five ways in. Pick the one that fits you.
Speaking. For anyone booking a speaker for a conference, an offsite or an internal event. I talk about AI adoption from the inside, out of the companies I have done it in and the workshops since. I held rooms for a living long before any of this.
An afternoon. For one person, or anyone who wants to see this work on one real task first. You pick something you do by hand today, and by the end of the afternoon it runs end to end on your real inputs. If it does not, you do not pay the second half.
A day. For someone who runs a team. A day built on their own work rather than a generic course, with a hands-on build after lunch, and everyone leaves with something they can switch on the next day.
The four-week pilot. For someone who runs a company and wants a number before committing further. Four weeks in three parts. We extract what your people already know, put it in a form your tools can read, and prove it on one real job against a baseline we take at the start. What comes out is yours.
The build. For someone who wants this in their own systems. A store of what your company already knows, in a form your AI tools can read from, and the first real job running on it end to end, with a person accountable for every word. It does not run your work on its own, and I will not build you one that does.
What people said after a workshop
Workshops
"This is not your mother's AI workshop! It is a clever, empowering angle that takes the narrative back from big tech and faceless corporate types optimizing everything."
"Loved Dan's enthusiasm, and that he came with 'controversial' statements and suggestions, which then he backed up with reasoning and evidence (and several years of experience). A true eye opener!"
"It was a fun and mind blowing workshop, with lots of real life examples which helped me to totally change my perspective of what AI is really capable of."
"This is one of the most sensible and ingenious approaches to collaborating with AI"
"With humor and wit, Dan lifts the fog of AI and brings clarity to what it can do for you"
"Invest in it, it's worth it!"
Two of these are running now
At reMarkable I built the store of what the company knew, in a form its AI tools could read from. By the time I left they were building it into the company properly and I was in the repository approving the work. Building it was the easy half. Getting a company of people to actually use it was the work, and that is the part I care about.
Hesketh & Son is the family firm, named for my father Philip. It finds tenders and grants, and the registers that quietly decide what a company may bid for. It runs on a system I built that reads the public record and drafts against what a scorer is really looking for, with a person accountable for every word.
Who you are
writing to
I am British, I live in Oslo, and most of my career has been inside Nordic technology companies: Huddly, Jabra, two years at reMarkable. Before that I was chief executive and co-founder of Upwave, which I helped rebuild and rename after it nearly went under.
Persuasion is the family trade. My father, Philip Hesketh, spent his career on the psychology of persuasion and wrote two bestselling books about it.
Before any of that I was a musician, four albums with Sony. It is the least relevant thing on my CV and quietly the most useful, because it is where I learned to hold a room.
Start with an email
Write to me about whatever brought you here. Plenty of people are not sure what they want yet and would rather think out loud with someone who has done this from the inside. Some have no staff to roll anything out to and want to know whether any of this is meant for them.
I read them myself and I answer properly, including when the honest answer is that I am not the right person for it.
The afternoon, the day, the pilot and the build each have a fixed fee, agreed in writing before they start, half on booking and half at the end. An NDA comes before I see any of your work, and nothing recurring happens unless you ask.