Human World RulesSee how people before you handled it.
Contents All

Learning and growth

Garry Kasparov

After losing to the machine

Garry Kasparov (1963– · Baku) — After losing to the machine.

Learning and growthModernAfter losing to the machine

What actually happened?

The 1996 match is usually left out, and it matters. He won it 4–2, losing only the first game — the first time a machine had beaten a world champion under tournament conditions, which at the time was itself the headline. He then had a year to prepare, and IBM had a year to rebuild. What he faced in 1997 was not the machine he had beaten. That is the ordinary shape of it: you win the first round against a tool, conclude the threat was overstated, and meet a different thing the following year.

Strike out what is already lost

The day machines won on depth of calculation, that track closed. Training on a closed track has no exit however hard you train. The first move isn't redoubling effort; it's removing the lost item from your own list of what you are worth.

Change the rules, not the effort

He didn't train harder and he didn't leave chess. A year later he staged a tournament where each side brought a computer to the board. When a tool matches an ability, the valuable position moves to how the tool is used — and someone has to define that position first.

The appeal and the next move cost the same hours

He spent years arguing the match hadn't been fair, and withdrew all of it in 2017. Proving the loss didn't count and working out what to train next come out of the same person in the same stretch of time.

How do I use it today?

When a core skill in your field gets matched by a tool, don't train harder first. Ask which item is now clearly lost, strike it out, and see what's left that the tool can't replace. And if you notice yourself repeating that it isn't actually any good, count how many times this month — then count what new thing you practised in the same month.

Deep read

After losing to it, he changed the rulesLet people bring machines to the boardStop competing on who calculates fasterThe winner wasn't the best human or the best machineIt was the pair with the best processTwo amateurs beat the grandmastersHe spent years arguing the match didn't countConceding is cheaper than appealingIBM dismantled Deep Blue. He had no opponent left

Read alongside

Norbert WienerOne built the thing and warned the people it would displace; the other was displaced and redesigned the contest.

Further

Norbert Wiener John Boyd

Lines to keep

Calculation was already lost, so stop competing on calculation.

The first step isn't more effort. It's striking out what's lost.

Whatever replaced you does not need your agreement.