Key takeaways

A three-month trial found that routine AI use improved work quality across the board.

Senior lawyers who used AI for 90 days showed stronger judgment.

Junior lawyers showed no average skill gain.

Their scores split into more strong scores and more low scores.

The source says AI can help or hinder learning, depending on how it is used.

Better output does not always mean better learning

A three-month randomized controlled trial with practicing patent attorneys found a mixed pattern. Routine AI use lifted work quality across the board. But the effect on learning depended on seniority.

Senior lawyers who used AI for 90 days showed stronger judgment than those who did not. Junior lawyers did not show an average skill gain. Their scores split in two directions, with more strong scores and more low scores.

The source frames judgment as the skill that separates senior professionals from junior peers. It also notes that judgment usually develops through thousands of hours of on-the-job learning. That learning often comes from routine work and guided supervision.

How AI changes on-the-job learning

The source argues that AI is already changing how people learn at work. It points to studies in radiology, business problem-solving, job-seeker writing, and legal education. In those settings, thoughtfully embedded AI tools helped less-experienced workers improve independent performance.

The source also points to other experiments with software engineers, management consultants, high school students, and clinicians. In those cases, AI acted like a temporary exoskeleton. It improved immediate output, but the gains often did not persist after the tool was removed.

This creates a central tension. AI can raise performance now. It can also change the path by which expertise develops. The source does not claim one outcome always wins.

What the study design tried to capture

The source says understanding AI's effect on expertise requires three elements that rarely appear together. First, AI must be used for months inside ordinary professional work. Second, researchers need a credible test of unassisted judgment when AI is unavailable. Third, domain experts must grade the work blindly.

That design matters because short tests can miss longer learning effects. A tool may improve output quickly while leaving deeper skill unchanged. It may also support learning in some roles and not in others.

The patent attorney trial is presented as an example of this longer view. It does not suggest that AI has the same effect for every worker. It suggests that seniority changes the result.

Why the findings matter

The source asks a simple question: does better work always mean better workers? Its answer is no, not necessarily. Better output and better judgment are related, but they are not identical.

For senior professionals, AI use may reinforce judgment. For junior workers, the picture is less clear. The source reports no average gain, but it does report a wider spread of outcomes.

That spread is important. It suggests that AI may not affect all learners in the same way. Some may benefit more than others. Others may rely on the tool without building the underlying skill.

Morocco relevance

The source reports no Morocco-specific facts. A general lesson for readers is that AI adoption should be judged by both output and learning, not output alone.

Bottom line

The source presents AI as both a productivity tool and a learning variable. In the patent attorney trial, it improved work quality, but its effect on judgment depended on seniority. The broader lesson is that better immediate work does not automatically create better long-term workers.