AI Compresses Experience.

I have watched three waves of technology arrive with the same promise and the same fear. The early internet was supposed to make expertise worthless because everyone would have access to everything. Machine learning was supposed to make judgment worthless because the algorithm would decide. Now we hear that AI will make experience worthless. Each time, the fear was aimed at the wrong thing.

I built neural nets and intelligent agents in the 1990s. The math was there. What was missing was judgment about where to point it. That is still the scarce input. A model can draft, summarize and simulate. It cannot know which customer matters, which risk is fatal, or which board member needs to hear it first.

Where the time goes

Think of Boyd’s loop: Observe, Orient, Decide, Act. For most of my career, the slowest parts were Observe and the early work of Orient. Gathering information took weeks. Building the analysis took more weeks. By the time a leader had a clean picture, the picture was often stale.

AI collapses those phases. The research that took a team two weeks now takes an afternoon. The first draft of the analysis appears in minutes. The leader spends far less time assembling the picture and far more time orienting honestly and deciding well. The loop runs faster without anyone working longer hours. That is compression: years of experience, applied at a tempo that used to be impossible.

What the research shows

Erik Brynjolfsson, Danielle Li and Lindsey Raymond studied the rollout of a generative AI assistant to 5,179 customer support agents. Issues resolved per hour rose about 14 percent on average and 34 percent for novice and low-skilled workers, with minimal effect on the most experienced. The tool did not invent that expertise. It carried what the best people already knew to everyone else, faster. Experience was the input the compression ran on.

That is why your experience is not obsolete. A machine can give a young founder a hundred variations of a pricing strategy in a minute. It cannot tell her which one her biggest customer will quietly resent. Someone who has watched that happen can. The danger is not that the tools replace you. The danger is that fear keeps you from picking them up.

Compress your own loop

1. Time your last decisionWrite down how long a recent major decision took from first signal to action. Mark which days went to gathering and which went to judging.
2. Hand off the gatheringGive the slowest observe task to an AI tool this week: the research, the summary, the first pass at the analysis.
3. Spend the time on orientationUse the hours you get back to test your read of the situation with the people closest to the problem.
4. Write what the tool cannot knowFor each big decision, list what only experience tells you: which customer matters, which risk is fatal, who needs to hear it first.