Nathaniel Whittemore assembles roughly 41 statistics from named surveys, spending data, workplace reports, and public-opinion research to show how uneven AI adoption remains. A Gallup poll puts workplace AI use above half of US workers, while several enterprise surveys report meaningful value without equally established or spend-adjusted return on investment.
The business data highlights a widening gap between average and leading adopters. Reported token costs are forcing companies to rethink plans even though many do not meter usage, and card-spending data shows a large difference between median buyers and the top one percent. The same data suggests Anthropic edged past OpenAI among surveyed business customers, while the episode notes that open-weight, fine-tuned, and routed multi-model strategies are harder to track.
Workplace behavior is similarly mixed. The cited studies report rising resistance to agents, stigma around disclosing AI use, policy-violating shadow use, more cross-occupation work, significant time spent supervising agents, and a rapid rise in AI-generated code. Hiring figures point in different directions: some employers reduce entry-level work or prefer AI investment, while heavy adopters and software-developer postings show continued or increased hiring.
The final statistics broaden from work to social impact. Surveys report low trust in AI companies, concern about electricity prices and local data centers, limited investment in staff training, shifts in shopping and professional services, emotional relationships with AI, and widespread use among children without corresponding parental safety conversations. The episode treats these findings as signals from different samples and dates, not one unified causal study.
Watch on YouTube
