In a recent chess match during the Candidates tournament, a player – Hikaru – took more than 60 minutes to think on a single move. There were two possible moves that were playable at the position. After more than an hour thinking, Hikaru chose the wrong one and he eventually lost the game. After the game, the US grandmaster accused his seconds (his preparation team) of not having prepared him properly for that specific opening variation. This episodes sheds some interesting light on how AI is used and on the boundaries of that use.
You see, all chess grandmasters use computers to analyze openings before a game. The computers, with their immense calculation power – can show the best move in every position. However, the application of that knowledge by a human player on the live chessboard comes with some challenges. One cannot remember of all possible moves and all possible variations – even GMs have a limit. So players at the Candidates level analyze the most possible variations that they know the other player might play and concentrate on these. Even there though, knowing the best moves isn’t enough. You should truly understand the position. You must actually know why each move is best – not just memorize by heart. Because if the other player plays something slightly different you will be in trouble – as happened with Hikaru.
Quick camera change to business analysis. Interior of a meeting. Zoom in the business analyst talking. Trying to convince the stakeholders why a feature is not needed, even though it was requested. Knowing by heart some things the AI tools provided you as output the day before won’t do the trick. You should actually own the situation holistically. Know the details. Understand everything. Know the politics and the intricate details of the project. Have good relationships with people who affect the decisions. Provide alternatives that can work and when asked questions you weren’t prepared for, to be able to articulate true mastery of the subject at hand.
Because no matter how helpful AI tools can be, they will not be with you in the meeting. They will not sit face-to-face with the customer trying to convince him of something new. As in the live chessboard, in real life you are on your own. And you must convince people why you are worth their time and trust. And no matter how many pages of deep analysis the LLM models have produced the day before, today you will be alone. Once upon a time we were presented with Google. And suddenly all information was in our hand in seconds. But no one became a professor of nuclear physics because he could retrieve information on nuclear physics in a second. Now we have AI tools. But no one will be a good advisor because they can copy-paste information from ChatGPT.
On the live chessboard you are on your own.
Free to fail. Free to achieve miracles.
Based on the things that know and truly own…
Sources
BA.AI 
Leave a comment