From Rented to Owned Intelligence with Baseten

Why It Matters

Owned intelligence lets enterprises control AI costs and performance, accelerating adoption in high‑impact sectors like healthcare and finance while reducing reliance on opaque, pay‑per‑token services.

Key Takeaways

• Baseten enables companies to shift from rented to owned AI models.
• Owned intelligence gives control over model quality, performance, and costs.
• Enterprise adoption rising in healthcare and finance due to clear ROI.
• Baseten offers shared APIs, dedicated inference, and post‑training tools.
• Continual learning research bridges inference and training for long‑horizon agents.

Summary

The video introduces Baseten, an AI infrastructure firm that helps businesses move from "rented" AI—pay‑per‑token, shared‑endpoint models—to "owned" intelligence, where firms fine‑tune and host their own models, controlling quality, latency, and expenses. Baseten’s vision is a future populated by many specialized models, each managed through continual‑learning pipelines that keep performance aligned with evolving data. Key insights include a three‑tier product stack: shared open‑source APIs for quick entry, dedicated inference for custom SLAs, and a post‑training suite that bridges training and deployment. Adoption varies by company stage: early startups prototype with closed models, hyper‑growth AI firms blend open‑source and custom models, while enterprises—particularly in healthcare and finance—are beginning to invest heavily as ROI becomes evident. The discussion cites real‑world examples such as Intercom, Decagon, Cursor, and healthcare pilots, illustrating how post‑trained models improve outcomes. Baseten’s recent acquisition of Parsed adds tooling for continual learning, enabling models to evolve in‑context during long‑horizon, agentic tasks. The company positions itself as a customer‑centric inference platform rather than a pure research lab. Implications are significant: businesses gain cost predictability, regulatory compliance, and performance guarantees while avoiding vendor lock‑in. As AI workloads shift toward agentic, tool‑rich workflows, Baseten’s expanding inference stack—combining reliable runtimes with higher‑level primitives—could become a critical layer for scaling next‑generation AI applications.

Original Description

Greylock Change Agents is a speaker series that explores the cutting edge of agentic AI. In this session, Greylock Partners' Corinne Riley interviews Tuhin Srivastava, CEO and co-founder of Baseten.

During the conversation, Tuhin explains why the fastest-growing AI companies are moving from rented to owned intelligence, post-training their own models to control quality, performance, and cost. They explore how the inference stack is evolving to support long-horizon agentic workloads, what Baseten's acquisition of Parsed means for the future of continual learning, and how Baseten's distributed multi-cloud architecture turned an early enterprise constraint into a lasting competitive advantage. They close on GPU capacity, frontier hardware, and how to build a business in a market that no one can fully predict.

Tuhin Srivastava, CEO and Co-Founder, Baseten → https://www.baseten.co/

Corinne Riley, Partner, Greylock → https://greylock.com/

More Change Agents episodes → https://www.youtube.com/playlist?list=PLnsTB8Q5VgnVh39-ENR9F0UYE0QyoPKme

0:00 - Intro and what Baseten does

1:24 - From rented to owned intelligence

4:46 - The AI adoption curve across company stages

7:25 - The Parsed acquisition and continual learning

10:29 - Agents and the evolving inference stack

15:03 - GPU capacity and multi-cloud infrastructure

21:38 - Frontier hardware and staying ahead of the market

#ChangeAgents #AIInfrastructure #Baseten