Reports on the threat that generative AI poses to knowledge workers usually focus what might be called “structured creativity”: tasks like copywriting or computer programming where outputs are clearly specified through things like creative briefs or system requirements. One group that generally doesn’t consider itself at high risk are the analysts and consultants who write the reports themselves. After all, their work requires insights and judgements that are beyond the current capabilities of genAI, which ultimately does nothing more than play back the training data it's fed. Case in point: This widely reported BCG study found that genAI tools substantially improved performance of consultants at product planning, with the greatest improvements among below-average performers. A second finding of the same study was less widely noted: AI depressed performance at generating insights from interviews and raw data.…