Why Does the Gap Widen Instead of Closing in the AI Era?
In a study analyzing how closely professional Go players' moves matched AI's, the player with the highest match rate was Shin Jinseo (9-dan) at 37.6%, with a 95.7% win rate. The player with the lowest match rate sat at 22.6%, with a win rate of just 27.3%. This relationship is no coincidence: broadly, the more closely a player matched AI, the higher the win rate — a proportional trend. Even though AI, the answer key, opened up equally to everyone, the closer a player was to AI, the stronger they became.
A documentary summed up the phenomenon this way: "In the age of AI, rather than everyone leveling up, the gap actually widens." The reason is clear. Top-ranked players understand and use AI more deeply than lower-ranked ones. In the past, even when skill gaps existed, they widened slowly; now, with results pouring out through AI, the gap widens fast. Hand the same tool to both, and the output diverges — and even the speed of divergence has accelerated.
Curiously, the opposite sentiment also exists. Master Cho Hunhyun says that after AI arrived, "individual style disappeared — you and I both have no choice but to play what AI recommends." It looks as if the tool erased personality. But what actually happened was the reverse. The moment everyone held the same answer key, a new gap opened up over who could digest that answer key better.
AI Is Only an Answer Key, and You Cannot Study From the Answer Key Alone
One researcher compares today's AI to "an answer key that tells you the correct answer." The problem is that you cannot actually study from an answer key alone. From beginner up to around amateur 5-dan, a player cannot even grasp why the move AI declares correct is correct. If you see the answer but do not understand it, the answer key is just paper.
That is why the people who truly use AI well are professionals or top-tier players. According to one study, the higher a player's level, the greater the improvement they can expect through AI. The answer key opened equally to everyone, but the ability to digest it as one's own concentrates in those who already have skill. So the gap does not shrink — it deepens.
The Essence of AX Is Not the Tool but the Work Flow
If you understand AX as merely adopting a tool, you cannot explain this gap. In Go, top and lower-ranked players alike use the same top-tier AI engines, like KataGo or Golaxy. The programs are identical, yet results diverge — because the ability to interpret AI's moves and apply them to one's own board differs from person to person.
Where does this difference show up? Through the first 50 moves, mostly memorized joseki, anyone's match rate with AI is high. The game is decided afterward, in the midgame, where you can only apply a move if you understand the flow behind why AI plays it. The difference between playing from memory and knowing the flow opens up exactly here.
Organizations are no different. The core question of AX is not 'which tool do we use' but 'how do we redraw the flow of this work.' Only those who know why an approval step exists, where work gets stuck, and what to automate so the organization actually gets lighter can apply AI to their own board.
Why Might People in Their 50s Be a Step Ahead?
People say the AI era is one of the rare events where everyone starts from the same line. From the standpoint of tool operation, that is true. But shift to the standpoint of work flow, and the starting line is not the same. The person who already knows the flow in their bones is usually the one who has worked long — someone in their 50s.
Why the approval chain is structured the way it is, which department creates the bottleneck, where exception handling keeps repeating — these are instincts you cannot learn from a prompt. If AX ultimately means redesigning this flow, then the person who knows the flow ends up on the upper side of the gap when they pick up the tool. The starting line may look the same, but they may already be a step ahead.
There is a condition attached, though. Being in your 50s is not an automatic advantage; you must be willing to translate the flow you know into the language of AI. If you know the flow but refuse the tool, that knowledge stays locked inside the organization and vanishes. Conversely, the moment flow meets tool, twenty years of experience is amplified all at once.
What Does It Take to Survive AX?
The opposite case is just as clear. For someone who has only processed simple tasks quickly without knowing the flow, AI becomes the tool that replaces their job first — because in raw processing speed, AI has already surpassed humans.
There is an extreme example. At one professional-admission tournament, an aspirant who tried to cheat using an AI program was sentenced to a year in prison. If you copy only AI's results without understanding the flow, you may look ahead for a moment, but it eventually surfaces and collapses. The tool cannot replace the flow.
So the condition for surviving AX converges on one thing: not the ability to clear tasks, but the ability to quickly read and redraw the flow of work. Just as Shin Jinseo took AI as a teacher and ended up on the upper side of the gap, the speed at which you learn the flow is what decides your place.