There is a moment every accountant eventually experiences when they realise that the machine has just done in thirty seconds what took them three hours. This is usually around 10:14am. Five minutes earlier they were happily explaining to a client why the accounts needed "a little more work". Now they're staring at a screen while an AI has reconciled the bank, categorised the transactions, drafted the journals, flagged the anomalies and produced a beautifully formatted set of accounts. The accountant refreshes the page. It is still there. The numbers are right.
This is the first problem.
For centuries, accountancy operated on a wonderfully simple principle: if something was complicated, give it to an accountant. The accountant would then spend several days making it look complicated in a different way. There would be spreadsheets. There would be reconciliations. There would be a meeting. There would be another spreadsheet. Someone would send an email saying, "Just chasing this." Eventually, three weeks later, the accountant would produce a PDF containing 47 pages of numbers nobody understood. Everyone would feel that something important had happened.
Then AI arrived.
Suddenly the computer didn't need three weeks. It needed twelve seconds.
At first, accountants weren't worried. This was automation, they explained. Automation doesn't replace accountants. It merely removes the repetitive work. This was comforting because almost all of accountancy was repetitive work. But nobody wanted to say that part out loud.
So the profession moved on to the next argument.
"AI can't understand the client."
Perhaps. Although it turns out the client mostly wants their tax return done, their accounts filed and someone to tell them how much money they can take out of the business without HMRC becoming personally interested in their existence. AI can probably handle that.
Then came the favourite.
"AI can't exercise professional judgement."
This sounded considerably more reassuring. Professional judgement, after all, is a very impressive phrase. It sounds like something that happens in a walnut-panelled room while a senior partner looks thoughtfully over a pair of spectacles.
In practice, professional judgement occasionally means staring at a £3,842 transaction for eleven minutes and deciding it probably belongs in "repairs and maintenance".
AI has now started doing this too.
It is becoming uncomfortable.
The next stage was denial. Accountants began explaining that AI would never replace experienced professionals because accounting was fundamentally about relationships. This was true. Unfortunately, the relationship was often between the accountant and an email entitled "URGENT – YEAR END ACCOUNTS", sent at 4:57pm on a Friday. The client would then disappear for six weeks. The accountant would send four increasingly desperate follow-ups. The client would eventually reply, "Sorry, only just seen this."
AI has no problem with this.
AI doesn't resent you. AI doesn't need a lunch break. AI doesn't go on holiday. AI doesn't have a dentist appointment at 3:30. And, most importantly, AI doesn't spend twenty minutes explaining why it cannot work because someone has uploaded a bank statement upside down.
The profession then discovered something even worse.
The clients were starting to notice.
For years, accountants had been able to charge respectable amounts for work that was difficult for ordinary people to understand. This wasn't necessarily dishonest. It was simply the natural economics of expertise. You know how to do something. I don't. Therefore, I give you money. A perfectly reasonable arrangement.
Until I discover that the software can do it for £29 a month.
Suddenly the question changes.
Not, "How much does my accountant charge?"
But, "Why am I paying my accountant to do something my laptop does before I've finished making coffee?"
This is not a question the profession has particularly enjoyed answering.
There are still areas where humans matter. Complex tax. Judgement. Advisory work. Strategy. Ethics. Relationships. The genuinely difficult stuff.
Unfortunately, there is a small problem.
Most accountants did not spend their Tuesday afternoon doing genuinely difficult stuff.
They were chasing missing invoices.
Again.
By 2030, the average junior accountant may look back at the old profession with the same fascination we now reserve for people who used to calculate navigation using stars. "Wait," the teenager will ask. "You manually reconciled bank transactions?"
"Yes."
"All of them?"
"Yes."
"Why?"
The accountant will stare into the middle distance.
"Because that's how we did it."
There will be museums. There will be an old Excel spreadsheet displayed behind glass. Visitors will press a button and watch a pivot table calculate. Children will laugh.
"Was this before AI?"
"Yes."
"How did they survive?"
Nobody will know.
Eventually, the final accountant will sit alone in an office. The lights will be off. The printer will be unplugged. The filing cabinets will have been sold. On the desk will be a calculator, a green visor and a single piece of paper containing a handwritten trial balance.
The accountant will look at it.
For old times' sake, they will reach for a pencil.
Then the software will quietly reconcile it in the background.
A notification will appear.
Accounts complete.
The accountant will sigh.
"Well, that's that then."
And somewhere in the cloud, without qualifications, without a practising certificate and without ever attending an audit seminar at a depressing hotel near Birmingham, the machine will have finished another year's accounts.
The invoice will be generated automatically. The client will pay it automatically. The records will be filed automatically. The tax return will be submitted automatically.
And the accountant?
They'll update their LinkedIn profile.
Former Accountant | AI Transformation Consultant | Helping Businesses Navigate the Future
Because the last great skill accountancy will retain is explaining why the thing that replaced you is actually an exciting opportunity.

Reader Comments
Leave a Comment or a Book Review
We would love to hear your thoughts on this chapter, or even a complete book review.