Mark Cuban recently reposted a career argument that has circulated widely in the AI era: AI will not take your job, but someone who is better at using AI will. His post on X drew nearly 2,500 likes.
Ethan Mollick, a professor at the Wharton School of the University of Pennsylvania, replied that the advice is "not as correct as it used to be." Mollick, who studies AI’s impact on work and organizations and has written on AI collaboration, laid out three reasons. His response collected nearly 600 likes and more than 30,000 views, fueling discussion in the AI community.
Mollick’s three objections
First, Mollick argued that "being good at using AI" is no longer a clearly defined skill. As AI products become easier to use, it is less clear what durable capability sits behind the idea of being better than others at using them.
Second, he said the constraint on AI gains is increasingly organizational rather than individual. In that framing, the biggest obstacles are internal rules and workflows, not whether a single employee knows how to use the tool.
Third, Mollick said the effect of AI on employment remains uncertain, but the disruption could land on entire job categories rather than on a narrow contest between coworkers over who uses AI more effectively.
From personal competition to structural limits
Cuban’s formulation places the relationship between AI and work inside a competition between individuals: learn the tools earlier, use them better, and you improve your chances of staying employed. It is an easy idea to understand, and one that can motivate people to learn new tools.
Mollick’s reply shifts the focus upward. If AI tools continue to lower the barrier to use, time spent mastering specific operating techniques today may matter less once interfaces improve. And even if employees know how to use AI well, companies may still fail to capture much value if approval processes, data access rules, and work allocation do not change with it. If AI ends up replacing a meaningful share of a given type of work, individual effort alone may not offset competition inside a shrinking market.
A cited example and the counterargument
The article points to an earlier report from Chain News that cited research by AI talent platform Mercor. According to that research, AI had moved ahead of junior accountants in accounting work after still lagging 18 months earlier. The example was presented as a sign that some entry-level roles may be facing broader pressure.
At the same time, the piece notes that Mollick’s three points were arguments, not data-backed findings in that post. It also says the claim that organizations are the main bottleneck may fit large companies better than freelancers, startups, or small businesses, where an individual’s ability to use AI can still directly shape output and competitiveness.
The article adds another point: even if AI tools become easier to use, the ability to decide what AI should do and to judge the quality of the output may not lose value at the same pace. From that angle, Cuban and Mollick may not be in full conflict. Individual skill may still be necessary, but no longer sufficient on its own.
What the debate suggests for workers
ABMedia says the discussion offers two reminders for office workers in Taiwan. Learning AI tools still matters, but stopping at operating tips is not enough; judgment and domain knowledge matter more. It also argues that personal effort has limits, and that whether a company changes its systems and embeds AI into actual workflows is what determines whether productivity improves.
The article also notes that Mollick recently said AI agents will enter enterprise customer service in large numbers. As AI moves from the tool layer into the structure of work itself, the familiar line that knowing how to use AI will keep someone safe from displacement looks less complete than it once did.