Harvard Business Review has published two articles warning that companies aggressively adopting generative AI are experiencing 'knowledge decay' — a feedback loop where low-quality AI output ('workslop') erodes organizational decision-making, trust, and institutional memory. A BetterUp-Stanford survey of 1,150 workers found 41% received workslop in a given month, costing roughly $186 per worker monthly, or $9M annually for a 10,000-person company. Social costs compound the financial ones: recipients view senders as less trustworthy and capable. Broader data from MIT Media Lab (95% of orgs saw no measurable ROI on AI) and Goldman Sachs (no economy-wide productivity link to AI adoption) reinforce the concern. HBR's proposed fix — adding human verification layers around AI output — ironically consumes the labor AI was meant to replace. The articles distinguish between targeted proprietary AI use and indiscriminate public LLM adoption, arguing the latter produces generic, error-prone content that degrades faster than headcount savings justify.

5m read timeFrom thenextweb.com
Post cover image
233 Impressions