In 1997, a machine finally beat the best chess player alive. IBM’s Deep Blue defeated Garry Kasparov in their 1997 rematch, and most people read it as the end of an era — the moment the computers won and the humans went home.
Kasparov read it differently. Instead of retreating, he asked a stranger question: what if you stop making it human versus machine, and start making it human plus machine? He called the format advanced chess, and several years later, a 2005 online freestyle tournament gave the clearest answer.
The winners weren’t grandmasters. They weren’t supercomputers either. They were two amateurs running three ordinary computers — players who would have been crushed by any professional in a normal game. Yet paired with machines and a sharp process, they beat strong grandmasters and powerful chess engines.
The lesson often attributed to Kasparov became famous: a weak player with a machine and a better process can beat a stronger machine alone — and even a stronger player using a machine with a weaker process. Process won. Leverage won.
I think about that result constantly, because it describes the opportunity sitting in front of every doctor right now. And almost nobody in medicine is taking it.
The second shift nobody pays you for
Here’s the part of medicine we don’t talk about. You finish your clinical shift, and then you start your second one. The unpaid one.
This is the portfolio that won’t write itself. It’s the reflections, the appraisal evidence, the audit, the teaching prep, the emails that breed overnight, the life admin you’ve been ignoring since your last set of nights. Moreover, if you’re trying to build anything outside medicine — a bit of income, a side project, a way out — that lands on the same exhausted pile.
Most doctors respond to this in one of two ways. Either they grind through it on willpower until they burn out, or they quietly let it slip and carry the low hum of guilt instead. Neither option buys back a single hour.
So we tell ourselves the only fix is more money — enough to drop a shift, hire help, or eventually walk away. That’s true, but it’s slow. The money lever takes years to pull. Meanwhile there’s a second lever most doctors never touch, and it works this week.
Two levers, not one
Freedom has two inputs, not one. The first is money — assets that buy back your time eventually. The second is leverage — systems that buy back your time now, before the money ever arrives.
Doctors are trained to respect the first and ignore the second. We’re a profession built on personal effort, after all. You worked harder, you revised longer, you stayed late, and it got you here. So effort feels virtuous and delegation feels like cheating.
But that instinct is exactly what kept those grandmasters losing to amateurs with laptops. The strong player trusted their own brain and refused the machine. The weak player trusted the process — and the process won.
AI may be the cheapest leverage most of us have ever had access to. For roughly the price of a coffee or two a week, depending on the tool, you have a tireless assistant who drafts, summarises, researches, and reformats on demand. The question is no longer whether it can help. Instead, the question is which of your unpaid jobs you’re still doing by hand out of habit.
The Centaur Stack: what I actually delegated
Let me be specific, because vague “use AI for productivity” advice is worthless. Here’s my real stack — what I hand over, and what it gives back. I call it the Centaur Stack, after Kasparov’s human-plus-machine players.
Portfolio and reflections. As an emergency medicine doctor going through the CESR route, my evidence burden is enormous. I used to lose entire evenings to written reflections. Now I dictate the raw case — what happened, what I felt, what I’d change — and AI shapes it into a structured reflective entry that I then edit and verify. The thinking stays mine. The formatting, the framework, the first draft: delegated. My estimate is that alone gives me back four or five hours most weeks.
Content and teaching. I run two writing channels and a teaching commitment on top of clinical work. One recorded voice note becomes an outline, a draft, and ten short posts — repurposed automatically rather than rebuilt from scratch each time. What used to be a weekend is now an afternoon.
Inbox and admin. My email gets triaged before I open it: drafted replies, flagged actions, deleted noise. Consequently I make decisions instead of doing data entry. Two hours a week, reliably, handed back.
Research and summarising. Before, a new guideline meant an hour of reading. Now I get a faithful summary in minutes, then go deep only where it matters clinically. The judgement is still mine — but the slog isn’t.
Add it up and, for me, it’s comfortably ten to twelve hours a week. That’s not a productivity hack. That’s a clinical shift’s worth of life, returned to me every single week, without earning a penny more.
Why most doctors get this wrong
If it’s this good, why isn’t every doctor doing it? Three reasons, and you’ll recognise all three.
First, the identity problem. Handing work to a machine feels like a threat to the thing we’re proud of — our competence. As a result we resist it, even when it’s freeing us to do the work only we can do.
Second, the trust problem. Doctors are quite rightly allergic to being wrong. AI gets things confidently wrong, so the instinct is to reject it wholesale. The better move is the centaur move: never trust it blindly, but never refuse the leverage either. You stay the verifier. It does the drafting.
Third, the input problem. Most doctors try AI once, type a lazy one-line request, get mediocre output, and conclude it doesn’t work. Yet the quality of what comes out depends entirely on the quality of what goes in. Garbage prompt, garbage result. A clear brief, your real context, an example of what “good” looks like — and the output transforms.
The 3D filter: deciding what to hand over
You don’t need to automate your whole life this weekend. You need a filter. I use three questions — the 3D filter — on any recurring task.
Delegate: does this need my judgement, or just my time? If it only needs time, it’s a candidate to hand over. Draft: could a competent assistant produce a first version I then refine? If yes, never start from a blank page again. Decide: is the real value in the final decision rather than the production? If so, protect the decision and delegate everything upstream of it.
Run your unpaid second shift through that filter. You’ll be surprised how little of it genuinely requires a doctor — and how much of it you’ve been doing by hand purely out of pride.
Where this actually leads
Here’s the uncomfortable truth underneath all of this. The reason most doctors feel trapped isn’t only the money. It’s the time. You can’t think about an exit, a side income, or a different relationship with medicine when every spare hour is swallowed by admin.
Leverage breaks that loop first. Buy back ten hours a week and you suddenly have room to build the financial side — the planning, the second income, the freedom number that eventually lets you practise on your terms. The two levers compound. Time funds the thinking that builds the money that buys more time.
That’s the whole game. And it starts not with a windfall, but with a decision to stop doing by hand what a machine could draft for you tonight.
Your action steps this week
- Audit your second shift. Write down every recurring non-clinical task you did this week. Be honest about the hours.
- Run the 3D filter. Mark each task Delegate, Draft, or Decide. Anything that isn’t pure judgement is a candidate.
- Pick one task and hand it over. Start with reflections, email, or content — whichever steals the most time. Give the AI a proper brief, your real context, and an example of “good.”
- Stay the verifier. Edit and check everything. The judgement stays yours; the slog does not.
- Reinvest the hours deliberately. Don’t let the reclaimed time leak back into more shifts. Point it at the thing that builds your freedom.
For more topics on building a life of time and financial freedom, sign up for our weekly newsletter at www.building-out.com
This post is for educational purposes only and does not constitute financial advice. Always do your own research and, if needed, ask for advice from a qualified financial adviser regulated by the FCA.
Leave a Reply