When I qualified, an accountant’s value lay in knowing things the client could not easily find out. The rules, the treatments, the story behind the numbers. Knowing was the job, and it paid well for a long time.
I have watched that change more than once in 50 years. Ledgers to spreadsheets. Spreadsheets to the cloud. Now the cloud to AI. Each time, the work I was paid for moved somewhere new, and some firms moved with it while others stood still.
So a book called AI and the Great Value Migration was always going to get my attention.
Peter Thomson has been in business since 1972. He wrote most of it, with three chapters from Darryl James on putting AI to work and one from Rachel Groves on running a business almost single-handed with AI behind her. It is short. You can read it in an evening, which is rarer in a business book than it should be, and to its credit.
The argument is simple. When a technology makes something abundant, the value drains out of the old work and gathers around whatever stays scarce. Ice gave way to refrigeration, and the money moved to the people who kept things cold. Kodak owned film and missed the photograph. The pattern is older than any of us, and AI is the newest turn of it.
What AI makes abundant is knowledge. Answers, first drafts, summaries, research. The very things a professional used to be paid to hold.
His useful move is to ask a sharper question of the technology. What becomes more valuable because AI exists? His answer is what he calls the Human Premium: trust, judgement, accountability, and the nerve to tell a client to act and mean it. As information gets cheaper, those get dearer.
For anyone who advises for a living, the message is blunt. Expertise used to be the summit. Now it is the price of entry. Clients assume you are competent. What they pay for is your reading of what everyone can now find out, and the outcome you are willing to stand behind.
An honest caution. The core idea is old. Adrian Slywotzky wrote a book called Value Migration in 1996, and this is that idea dressed for the AI moment.
The last third earns the cover price on its own. A monthly habit for spotting where value is moving. A test for which tasks to hand to AI and which to keep on your own desk. A one-page method for teaching a process to a machine so it stops living only in your head. I have started using a version of that last one myself.
Thomson’s own one-line version is “Don’t follow technology. Follow the value.”
After 50 years of watching exactly that happen, I think he is right.
If the pace of AI has you watching the tools and wondering where you fit, this is a calm and useful place to start. There are no prompts in it and no tool tips, which is deliberate. It is a book about where to stand while everyone else is staring at the screen.
I’ve applied the book’s questions to UK accountancy, from the side of the owner who buys it, in The Goldmine in Your Ledger.
