AI Makes Tech Workers Faster, Not Better
Workers say AI speeds output while eroding thinking, unless they force themselves to practice judgment the hard way.
Lenny's Podcast: Product | Growth | Career · Why the AI’s honeymoon is ending (and tech workers are feeling it) | Noam SegalNearly every survey respondent says artificial intelligence makes him or her better at the job. In a survey of about 6,000 people in tech roles, 97.2% said AI made them better, and nearly half said much or extremely better, according to Noam Segal, who ran the survey and is also a certified coach.
But better does not mean what it sounds like. When Segal asked what people meant, the answer was quicker output, not higher quality work. Segal said AI has "lowered the floor of what's possible," letting people put out a much larger volume than before, whether writing software, drafting plans, doing research projects, or building early versions. The work arrives faster, yet workers do not say it is superior.
The reason lies in what happens between the machine and the mind. Workers described looking at the first answer from a model and going along with it, without pausing to check it or think it through. For that pattern, Segal borrows the phrase cognitive rot, which he said he had heard used elsewhere: the person steps back, judgment fades, and thinking gets less sharp over time.
People know it is happening, and Segal says knowing is not enough. As a coach, he argues workers have to act on that awareness. He said that because people will press the easy button when it is there, staying sharp takes what he described as "deliberate practice, deliberate thinking, deliberate judgment," not just noticing that skills are slipping.
Each shortcut has a cost to confidence. Segal said he once thought of himself as smart, then met newer models that made him feel less so. Solving a problem yourself builds belief in your ability, while handing it to a model lowers that baseline. In words Segal read from his team's report on the same survey, "productivity gains are real, but the quality of the work and the sharpness of the person producing it are taking a hit." The conversation then turned to a question about junior workers learning to code, write, or draft a plan from scratch rather than by asking a model.
Questions people ask
What exactly is cognitive rot?
It means gradually relying on AI answers without checking or improving them, so your own evaluation skills weaken. Like a muscle that moves less, judgment gets softer when AI does the first draft, the revision, and the decision every time.
How can I keep AI from dulling my thinking?
Do the task yourself first, then compare with AI, and force a choice about what is better and why. Coaches also recommend picking one or two high value tasks to practice deeply, and regularly writing or coding from scratch to keep the underlying skill alive.
Why are junior workers especially vulnerable?
They are most at risk because they may skip the stage where skill is built through repetition and mistakes. If a junior employee never writes basic code or a plan unaided, there is no foundation to judge whether the AI version is good, safe, or complete.