New study finds it 'very difficult' to measure actual financial impact of AI spending
Gary Marcus · rss · 2026-10-14
Gary Marcus cites a new study (via The Information) reporting it remains 'very difficult to find actual, materially measurable financial impacts' from corporate AI spending — a stark contrast to early-2023 claims that ChatGPT would 10x productivity, with trillions invested but little evidence of ROI at scale.
He also flags two industry signals:
- Early generative AI darling Character.AI has filed for bankruptcy;
- CoreWeave bonds have reached 15% yields, suggesting bankers have had enough.
More from Venture
- Exa hits $100M annualized revenue after 3 years of near-10x annual growth — Scobleizer · 2026-10-14
- The Structure Agents Need Isn't Technical — It's Operational — usehive · 2026-10-14
- Polymarket odds for a Chinese AI model hitting #1 by year-end slump to 11% — ZeroStateReflex · 2026-10-14
- A backyard "distributed data center" ADU startup for local LLM enthusiasts — dee_hw · 2026-10-14
- Mattermark creator launches StartupGraph: startup data API and MCP server for AI agents — Scobleizer · 2026-10-14
- WSJ survey: nearly 100% of firms worried or skeptical about AI, up 20+ points — GaryMarcus · 2026-10-14