According to Ethan Mollick on X, a widely shared influencer post misrepresented an AI creativity study; as reported by Mollick, the paper observed 61 participants and found no drop in creativity after 30 days, with the ChatGPT group remaining significantly higher at the end, citing the original X thread and linked article for evidence. According to Mollick, users should verify posts by pasting the claim and source link into a frontier model to check whether the post is supported by the evidence, noting that Grok does not read PDFs. The business takeaway, according to Mollick’s analysis, is that claims about generative AI harming creativity may be overstated, suggesting enterprises should rely on primary-source evaluation and model-assisted evidence checks before adjusting knowledge-work policies.

According to Ethan Mollick, a widely shared post misrepresented a creativity study on ChatGPT; as reported by Mollick citing the paper’s results, a 61-participant experiment found no decline in creativity after 30 days and the ChatGPT group remained significantly higher at the end of the period. According to Mollick’s summary of the study, the small sample size indicates the trial was underpowered, but reported statistics still showed sustained creativity gains for ChatGPT-assisted participants over time. As reported by Mollick, this challenges narratives of rapid skill atrophy from AI use and suggests business opportunities in structured, longer-horizon adoption of generative AI for ideation, marketing copy, and product concept generation, where month-long outcomes matter. According to Mollick’s interpretation of the authors’ data, organizations should prioritize measurement frameworks that track longitudinal creativity metrics and implement controlled rollouts with baseline and follow-up assessments to capture durable productivity and creative benefits.