Venture capitalist and All-In podcast co-host Chamath Palihapitiya says enterprises will pivot away from single-model AI dependencies within the next three years.
He argues that as AI model costs converge, the competitive advantage of being locked into any one AI provider will diminish significantly.
Palihapitiya wrote on X that businesses will increasingly turn to independent third-party software layers capable of working across multiple AI models simultaneously.
“As model costs converge, the next likely step will be the use of independent 3rd party harnesses/control planes (8090 Software Factory, Palantir, Nous’ Hermes etc),” Palihapitiya wrote.
He expects the shift to accelerate as companies “realize the importance of data/IP leakage,” making control over proprietary information a central enterprise priority.
Third-party harnesses sit atop different AI models and allow companies to switch between providers, use different models for different tasks, and avoid dependence on any single vendor.
Palantir Technologies (NASDAQ: PLTR) is positioned as a potential beneficiary, given that its AIP product architecture already supports models from OpenAI, Anthropic, Meta, Google, and SpaceX’s xAI, among others.
Palantir explicitly describes AIP as “model-agnostic,” and its Ontology operating layer allows enterprise clients to deploy different models for different workloads within a single unified platform.
That flexibility becomes increasingly valuable in a world where model performance and pricing converge, shifting competitive power toward the software layer sitting between enterprises and AI providers.
PLTR stock edged 0.5% lower in pre-market trade following Palihapitiya’s comments, with retail sentiment on Stocktwits slipping to “neutral” from “bullish” territory over the prior day.
Palantir does not have the field to itself, as Microsoft, Amazon’s AWS, and Google are also building out their own multi-model AI platforms targeting enterprise customers.
The broader debate over which companies stand to gain most from AI infrastructure maturity is intensifying, with control planes and data governance tools increasingly at the center of enterprise AI strategy.
