‘I’m a human being, you know . . . When I see something that is well beyond my understanding, I’m scared.’
– Gary Kasparov

Gary Kasparov is a force of nature. Born in 1963, he was a Russian chess grandmaster and world champion. From 1984 until his retirement from competitive chess in 2005, Kasparov was ranked world number 1 for a record 21 years and 3 months.
He was supremely gifted as a child, but as an adult whenever he walked into a room he was like a tornado, everyone would pin their backs against the wall. They were in awe of this man – sparks just bubbling out of his head.
He was also aware of his own power, and knew how intimidating he was. When he was playing chess against another player he would sit down, take his watch off and place it on the table. And then he would play. And when he was done toying with you, he’d pick up his watch and return it to his wrist.
Even if you hadn’t worked out that you’d lost, he would know that he’d beaten you. That was essentially a message to the room, ‘Right, it’s done.’ And you can either carry on humiliating yourself from that point on, or you can submit.
In 1997, a computer from IBM named Deep Blue was the first machine to beat this chess grandmaster. The program was very good at this point, but most people agree that Kasparov was still the better chess player. What the computer didn’t have was any fear and it wasn’t intimidated by Kasparov. And he wasn’t used to this. Suddenly, all of his tricks were useless.
Kasparov hadn’t played against a computer before and hadn’t worked out the psychology of how to do this.
So, the IBM researchers did something very clever.
The way the program worked was that it would search through all the different possibilities, so that if it was forced into a particularly tricky position, there would be many more calculations to do. This would take a lot longer for it to churn through them.
The researchers knew that Kasparov knew that if it was tricky, it would take a lot longer to come up with an answer and its next move. So what they did was to deliberately put some code into the program that would add on a random amount of time before it came up with an answer. So Kasparov then couldn’t tell whether Deep Blue was thinking hard, or not thinking at all.
Essentially, the IBM researchers psyched him out. There were points when the machine was waiting for ages and ages and Kasparov didn’t know about this extra little bit of time filler. In situations which seemed like a simple position and the next move was obvious, Kasparov would then wonder, ‘What has the machine spotted that I haven’t spotted?’ And so he started doubting himself and questioning himself. And he says, and others agreed, that that was what led to him making mistakes during the game and eventually losing the match.
The version of Deep Blue that defeated Kasparov in 1997 typically searched to a depth of six to eight moves and twenty or more moves in some situations.
It’s not just about the power and the might of computer algorithms versus humans. We have had tools that are super human for ages. Fork lifts are super human. But what happens with Artificial Intelligence is that they make us question ourselves.

(Source: Hannah Fry on ‘Snowcast’, with John Snow, 23 May 2023)