AI value shifts from model layer to data center and apps as harness beats models, analyst argues
RihardJarc · x · 2026-10-10
RihardJarc published an essay arguing AI value is moving from the model layer to the data center and application layer, with implications for $AMZN, $GOOGL, $META, $MSFT, semis and neoclouds.
Key points:
- The harness matters more than models for most knowledge work; one large token consumer says the bottleneck is context, not model quality
- Everyone is multi-model (average 5+ models) and open-source models are growing; transferable KV cache across models is the final step
- Application-layer margins are expanding
- Segment-by-segment outlook for hyperscalers, semis, model labs, neoclouds and apps
More from Venture
- Designer Signs and Gets Paid for 2 US Website Projects Delivered Entirely With Claude — Tegadesigns · 2026-10-10
- Orbio raises $1.2M to build rails for agents to provision their own compute — econoar · 2026-10-10
- dbskill turns 10.5k stars into seven-figure revenue: selling skills is e-commerce — oran_ge · 2026-10-10
- Lumentum's AI optical components sold out through early 2029, CEO says — shashib · 2026-10-10
- It's not an AI bubble, it's a SaaS bubble — mallow610 · 2026-10-10
- Ben Affleck reportedly sold his AI company to Netflix for $587M — AIFlow_ML · 2026-10-10