Six months after ChatGPT launched, most people using it were doing exactly one thing: writing emails faster. They were pasting in bullet points and asking for a professional version. They were getting it to draft responses they would have spent twenty minutes on before. Useful, sure. Transformative in the way the technology actually is? Not remotely. The gap between what most people were doing with it and what was actually possible was one of the widest in the history of any technology rollout.
The people building real advantages were doing something different. They were not using it as a faster word processor. They were using it as a thinking partner — feeding it complex problems, arguing with its responses, using it to pressure-test assumptions they had been carrying unchallenged for years. They were using it to compress timelines that previously required teams: research, synthesis, first drafts, competitive analysis, product spec, legal review. Not as a replacement for expertise, but as a lever that made their expertise dramatically more productive.
The founders who moved fastest in 2023 are already operating at a pace that will be difficult to compete with. Not because they have better ideas — ideas have never been the scarce resource. Because they are iterating faster. The product assumption that used to take three months to test now takes three weeks. The content strategy that required a team of four now requires one person with strong taste and good prompts. The gap compounds every quarter.
What most people missed about the technology is that the value is not in what it can do in isolation. It is in what it can do in combination with a person who knows what good looks like. An AI that can produce a competent analysis of any market is useless to someone who cannot recognize whether the analysis is correct. An AI that can write a persuasive email is useless to someone who does not know what persuasive looks like in their specific context. The tool amplifies the judgment of the person using it. That means the people with the most developed judgment get the most amplification.
The people who are not getting value from AI tools are usually the ones who are using them to outsource thinking rather than to augment it. They ask for a strategy and get a strategy and cannot evaluate whether it is right. They ask for an answer and accept the answer without understanding the reasoning. That is not leverage — it is the illusion of leverage, and it tends to produce mediocre outputs at high confidence, which is worse than no output at all.
The honest version of what changed in November 2022 is this: the productivity ceiling for any individual with good judgment and strong taste went up dramatically. If you have something worth saying, something worth building, or a problem worth solving, the constraint of your time and resources matters less than it did before. The people who understand that and are operating accordingly are pulling ahead. The people still using it to fix their email subject lines are not.
The next wave is not going to be more impressive demos. It is going to be the widening gap between the organizations and individuals who integrated this deeply into how they work and the ones who are still treating it as a novelty. That gap is already visible. In two years it will be obvious. The question is which side of it you are building toward.
