Third Party Evaluators, Fed Hikes Rates, Zuck Pushes Back On AI Slowdown | Jeremy Allaire, Tomasz Tunguz, William Layden, Justin Beroz, Eli Wachs, Sean McCarthy, Tom Mueller
September 16, 2026
AI Summary
5 min readThe Fed raised rates for the first time in three years, Mark Zuckerberg published a lengthy post defending Meta’s approach to AI safety that critics say mostly talked past the central debate, and Anthropic’s agreement to give third-party evaluators at METR badge-level access to its systems set off a broader argument about who should be trusted to audit frontier models. Those three developments dominated a packed episode that also covered the Commerce Department’s shutdown of an AI compute prediction market, Circle’s launch of a new blockchain-based operating system, a $308 million funding extension for Impulse Space, and a handful of startup product launches.
The Fed Hikes, and the Rate Outlook Shifts
The Federal Reserve raised its benchmark rate by a quarter point to a range of 3.75% to 4%, a decision the hosts called a sharp reversal from the cutting cycle last year. The move was unanimous. The vast majority of officials penciled in at least one more hike this year, and four now expect two additional increases — up from just one official at that level in the previous projections. In June, half the committee had expected the Fed to hold or cut rates.
Continue reading the full summary in the app — free to try.
Read Full Summary →Free • No credit card required
Never miss an episode of TBPN
Get every new episode summarized in your inbox — free, ~5 minutes to read.
No spam. Unsubscribe anytime.
What you'll learn
- 1 (00:00) **Show Open & Host Banter** - Hosts welcome the audience and discuss the format of the show, recorded in San Francisco.
- 2 (00:56) **The Debate Over Third-Party AI Evaluators** - The hosts dive into the controversy surrounding METR as a potential third-party evaluator for frontier AI labs.
- 3 (06:30) **Historical Precedent: Nuclear Regulation and AI** - The hosts explore the history of nuclear regulation, specifically the Atomic Energy Commission (AEC), as a model for AI governance.
- 4 (10:24) **The Moving Target of AI Safety** - The conversation shifts to the practical challenges of defining and measuring safety in rapidly evolving AI systems.
- 5 (13:15) **Fed Hikes Interest Rates for First Time in Three Years** - The hosts break down the Federal Reserve's decision to raise interest rates by a quarter point.
- 6 (17:10) **Silicon Valley's Reaction to Higher Rates** - The hosts discuss the mood in Silicon Valley regarding the return of higher interest rates.
- 7 (19:06) **Zuckerberg Pushes Back on AI Slowdown** - The hosts analyze Mark Zuckerberg's recent essay, arguing it misses the main point of the AI safety debate.
+ Full timestamped outline available in the app
Guests on this episode
Show Notes
- (00:52) - Third Party Evaluators
- (14:00) - Fed Hikes Rates
- (18:47) - Zuck Pushes Back On AI Slowdown
- (37:44) - Kalshi's AI Compute Market Closed
- (49:45) - Jeremy Allaire discusses Circle’s launch of Arc, a blockchain-based economic operating system designed for payments, capital markets, and trusted AI-agent transactions. The Circle co-founder, chairman, and CEO also explains how stablecoins such as USDC could reduce payment costs and expand global financial access, while addressing evolving U.S. cryptocurrency regulation.
- (58:29) - Timeline Reactions
- (01:05:21) - Tomasz Tunguz, founder and general partner at Theory Ventures, discusses AI regulation, liability, inference infrastructure, open-source models, and investment opportunities across the AI stack. He also examines soaring valuations for AI companies, agent economics and monetization, software acquisitions, and venture capital liquidity.
- (01:26:39) - William Layden discusses how Roon’s modular AI data centers convert surplus utility-scale solar power into computing capacity. He explains the technology’s rapid deployment, potential applications for wind and other renewable sources, and his belief that solar energy will power the future of computing.
- (01:35:25) - Justin Beroz, founder and CEO of Rainco, is a mechanical engineer and theoretical physicist developing mathematical tools to solve turbulence. Justin Beroz discusses the practical limits of the Navier–Stokes breakthrough, AI-generated proofs, and Rainco’s plans to commercialize more reliable fluid-simulation software and eventually turbulence-control hardware.
- (01:47:30) - Eli Wachs, founder of Footprint, discusses building an AI operating system that helps financial institutions detect, investigate, and prevent financial crime. He explains how Footprint uses AI agents, extensive data access, and case memory to investigate every transaction, counter increasingly sophisticated criminal networks, and reduce the multitrillion-dollar global impact of fraud.
- (01:58:31) - Sean McCarthy discusses his role as co-founder and CEO of Back Ops, an AI-native platform that automates supply-chain back-office operations. He explains how the company reduces operating costs, improves claims approval rates, integrates with systems ranging from spreadsheets to mainframes, and plans to use its $42 million Series B to expand across retail, manufacturing, and pharmaceuticals.
- (02:05:52) - Tom Mueller discusses Impulse Space’s $808 million Series D financing, its Mira and Helios spacecraft, and growing opportunities in orbital transportation and government launch programs. The founder and CEO also reflects on SpaceX’s engineering culture, AI adoption, spacecraft reliability, and the challenges of hiring experienced aerospace and software talent.
TBPN is made possible by:
Ramp - https://ramp.com
Public - https
More from this podcast
TBPN →