Are AI Foundation Models a Value Trap? History Says Apps Win
broodsugar · x · 2026-08-11
By drawing parallels to the history of the internet and crypto, the post questions current AI investment logic. The author points out that heavy investments in 'base layers' like telecoms and crypto L1s ultimately proved futile, as all the value accrued to applications.
Consequently, the author implies that the current trend of taking out a second mortgage to invest heavily in AI foundation models might be another bubble where value capture is misplaced.
More from Venture
- OpenAI Buys Back $7 Billion in Shares via Employee Tender Offer — shiringhaffary · 2026-08-11
- Applied Compute Raising Funds at ~$3B Valuation, Hits $50M ARR — steph_palazzolo · 2026-08-11
- Startup Hits $3B Valuation for RL Fine-Tuning Open Models — Hesamation · 2026-08-11
- Robinhood launches fund to invest in YC startups, lowering barrier to early-stage investing — bosmeny · 2026-08-11
- Open Source Catching Up Fast: Is Training Frontier Models Still Worth Millions? — 0xsachi · 2026-08-11
- YC Startup Prodigy Launches Quant Foundation Model, Claims 100%+ Live Returns — ycombinator · 2026-08-11