Does using an AI helper close the skill gap between a brand-new employee and a seasoned veteran?

This explores whether AI assistance helps beginners more than veterans, and whether that boost lasts. The short answer from the collection is yes, at least on the task in front of them. The more interesting answer is that "gain" may mean something different from what it seems.

The clearest evidence comes from two large workplace studies. In one, 5,172 customer support agents at a Fortune 500 company got about 15% more productive with an AI assistant. Most of that came from newer agents, who got faster and better at the same time Does AI assistance help less experienced workers most?. The most experienced agents sped up a little, but the quality of their work slipped slightly. That suggests AI suggestions can get in the way of people who already know a better answer. A separate experiment with 758 BCG consultants found the same pattern. Bottom-half performers gained the most from GPT-4, and the gap between weaker and stronger consultants shrank Does AI help lower-skilled workers catch up to top performers?. There is a catch, though: everyone in that study was required to use the AI. When using it is optional, the people who choose to adopt it may be the ones who were already ahead, and the leveling effect could flip.

The part you might not expect is what happens after the AI is taken away. Workers who did much better on content tasks with generative AI showed no improvement when they later did similar tasks on their own Does AI assistance help workers learn lasting skills?. Another note finds that the productivity gains in these studies come from tasks where workers already had the relevant skills. When people used AI to learn something new, the gains disappeared and their learning suffered When does AI actually boost worker productivity?. So the novice boost looks more like borrowed skill than built skill. The AI may be lending newcomers the know-how of top performers without passing it on to them. That matters a lot if the way newcomers become experts is by struggling through exactly the work the AI now handles.

Workers themselves see it more optimistically. Anthropic's Economic Index found that the people who hand the most work to Claude expect the best career outcomes and feel their skills are becoming more valuable Does delegating work to AI actually damage worker skills?. But that is self-reported optimism among Claude's own users, not measured skill. On the hiring side, the leveling idea shows up in a different form. Listing AI skills on a résumé partly offset the interview penalties that older candidates and those without a bachelor's degree usually face Can AI skills help older or less-educated job candidates?. Exposure to AI isn't spread evenly either. In female-dominated occupations it reaches all the way down to lower-paid, lower-skilled workers, who have fewer resources to adapt Does AI exposure hit low-wage workers harder in some fields?.

To sum up: on the job today, less experienced workers do gain the most. The open question is whether AI is turning them into experts or just covering for missing expertise. The collection leans toward covering. That makes "who benefits most" partly a question about how training gets designed, not only about the technology.