Contradictheory: Who to trust – computers or human intuition?

  • Living
  • Tuesday, 20 Nov 2018

Magnus Carlsen plays Fabiano Caruana at the 2018 World Chess Championship. Photo: Reuters

If chess is a hobby of yours, you will know that the World Chess Championship is now taking place in London. World No. 1 Magnus Carlsen is facing World No. 2 Fabiano Caruana – one on one, man on man – in a battle of wits and wills over chequerboard.

These modern-day gladiators pit against each other to calculate the best moves over the board all on their own – except that’s not quite true.

Each player has a team of assistants to help them prepare. These seconds, strong players in their own right, peruse the opponents’ past games to find potential weaknesses and strengths. And then they look for the “right” moves using computers.

Truth is, up to the first 20 moves of a game may have already been prepared, with each player hoping to spring a “surprise” move first. The role of computers in top-level chess is now commonplace, and people who play top-level chess really represent a hybrid of the best man and machine have to offer.

Computers not only help players but fans too. I don’t think most people would describe watching a chess game as “fun”, and even I would be pressed to sit still and watch a single game for hours – the exception being if it was a game I was playing.

On top of that, you have to understand how to play chess to appreciate and enjoy watching games. That’s why it helps to have expert commentators. Usually grandmasters in their own right, they give insight into what tactics and plans are possible from certain positions. If you’re lucky, they’re entertaining as well.

They also use technology, usually running chat rooms so that audiences can ask questions online.

One thing they don’t use computers for is to analyse games for commentary purposes. It’s strange because they respond to viewers in the chat rooms and those members of public use computers to help highlight moves. But the commentators themselves don’t run analysis programmes.

That said, these commentators do use computers in their post-game analyses, so it’s not that they are averse to them.

You see something similar in the corporate world. A study conducted of CEOs globally by management consultancy KPMG found that the majority of those surveyed chose to “overlook data-driven insights to follow intuition instead”.

The argument they make is that they sometimes find that what the computer says contradicts their own experience or intuition, and thus they are not confident about the accuracy of predictive analytics.

This result puzzles me somewhat. Is it saying that these leaders, presumably among the most informed people in society, don’t think results generated by computers are good enough – or is it that they don’t know?

I can accept the argument by chess commentators that they want to give a human perspective on a game being played live, to possibly peek into the mind of the best players in the world, so that audiences can empathise with the competitors.

In other words, you don’t want to be biased by the “right” answer too quickly because it is the journey, following down the path of imperfect choices, that make chess interesting.

Perhaps CEOs feel that computers can never completely capture that element that makes us human. What about creativity? Intuition?

I previously wrote about the Alpha Zero project, where a chess programme used a self-learning A.I. which seemed to make “intuitive” moves in games. So let’s look at other games, like poker.

Poker is a game seen to be very human because there is bluffing and guessing. Guess what? In Jan 2017, a programme named Libratus went up against four human professionals and beat them (although the humans still went home with US$20,000 as compensation).

I think the reason why humans prefer to trust themselves over computers is pride or even hubris. We want to look down on computers. We talk about the fear that they are taking over our jobs and way of life when, in fact, it is yet another element of change.

I think the real value in the ever-growing power of computing ability is to accept and embrace the change they are bringing. The regulation surrounding it should not be about what we fear but what we can do.

It reminds me of what one speaker said about GDPR, the European laws related to data privacy, that it’s not about what companies can’t do in their businesses, it’s about what they should do if they want to do it properly.

You are beginning to see this in terms of how computers are used in chess.

As mentioned, grandmasters accept it as commonplace for preparation and analysis, and I know some correspondence players (those who play one move per day over the Internet) who think it’s OK to use computers to “suggest” moves in the middle and end game.

On top of that, to discourage over-preparation, there are variations such as chess 720 where the starting positions of some pieces are randomised.

Perhaps the ultimate chess tournaments are played exclusively by computers. They’re not popular to watch, but there are those who select games for analysis. Sometimes it’s to feature a new idea in the opening, but other times it’s to highlight a computer doing something it’s not supposed to – playing like a human.

Logic is the antithesis of emotion but mathematecian-turned-scriptwriter Dzof Azmi’s theory is that people need both to make sense of life’s vagaries and contradictions. Write to Dzof at

Article type: metered
User Type: anonymous web
User Status:
Campaign ID: 1
Cxense type: free
User access status: 3

Across The Star Online