Chris Taylor, CEO of Ode, says that a slowdown at the AI frontier would have minimal impact on large companies because they are not operating at that frontier. Ode is a $1.5 billion joint venture formed by Anthropic, Blackstone, Hellman & Friedman and other investors, focused on implementing AI inside big enterprises.
"Even if the technology stops advancing — turn ‘pacing the frontier’ into ‘halt the frontier’ — we have a decade-plus of diffusion of benefits into the economy with the technology as it exists today," Taylor told Semafor. He argued that existing models already provide substantial capabilities that will continue to be deployed across businesses.
Recent data and corporate choices
According to data from Ramp, corporate use of the most cutting‑edge models has fallen in recent weeks, as companies recognize that many practical tasks—such as onboarding customers and iterating prototype tweaks—can be accomplished with older, cheaper models. Those models also carry fewer of the broad, systemic risks that are sometimes discussed at the frontier.
Taylor also noted that even Microsoft last month switched many employees to less‑powerful models.
"As the capabilities improve, it does unlock more, but the amount of possibility that already exists with yesterday’s models is staggering," Taylor said. He does not see these trends slowing AI adoption among enterprises.
Why this matters
A reduction in frontier usage and a shift toward older models implies that the economic diffusion of AI benefits will continue over many years, possibly more than a decade, even without further rapid advances at the frontier. At the same time, the significant utility of current models means companies can prioritize cost efficiency and risk management while continuing to integrate AI into operations.
Reporting: Liz Hoffman (original reporting)
Sources cited: Chris Taylor (Ode), Ramp data, reports of Microsoft’s internal model switches, Semafor interview.
