Sep 14, 2026 · 52 min
Listen from 49:53
Listen at 49:53
AI leaders clash over who should control the slowdown
AI’s Biggest Players Debate a Frontier Slowdown
The debate links frontier-model safety and oversight to competition, infrastructure demand, regulation, markets, and the future of AI-powered creation.
- 1Frontier AI leaders are calling for more deliberate development while independent evaluation and regulation struggle to keep pace.
- 2Investors weigh whether slower model progress would weaken demand for chips, cloud infrastructure, and AI-linked stocks.
- 3Roblox and Cohere offer competing visions of AI’s next phase: faster creation and broader governance beyond dominant labs.
Don't miss
Aidan Gomez makes the case that AI safety standards should not become a cartel controlled by a few dominant frontier labs.
The brief
Dario Amodei, Sam Altman, and Elon Musk have called for a more deliberate pace in frontier AI, but coordinating labs, governments, and evaluators remains difficult.
Investors debate whether slower frontier development would reduce demand for chips, memory, and cloud infrastructure, while older and open-source models could sustain enterprise adoption.
Anthropic’s proposal to give independent evaluators near-employee-level access raises a practical question: can oversight become a credible industry service without conflicts of interest?
Greg Brockman argues for objective evaluations amid anxiety about advanced AI, while Aidan Gomez warns that a few frontier labs should not control safety standards.
David Baszucki describes Roblox Build and new Studio features that could accelerate 2D and 3D game creation, while expanding experiences beyond the platform.
The episode’s central tension is whether pacing means a genuine slowdown or simply more deliberate development—and who has the legitimacy to decide.
What was said on this episode
10 statements · 3 positive · 6 negative · 1 neutral
Slowing frontier development spreads AI infrastructure demand over time rather than reducing total demand.
“slowing down the frontier doesn't actually lead to a slowdown of the total demand curve, but just basically spreads the curve out. over a longer period of time”
Listen at 8:58
Uniform AI regulation gives incumbent frontier labs a strategic advantage over slower competitors.
“the companies who are already ahead, the Frontier Labs and OpenAI, as well as Anthropic, Gemini, all of them, they have a strategic advantage”
Listen at 10:05
AI pacing could give Anthropic and OpenAI a clearer IPO narrative.
“the pacing argument actually helps them get a clearer IPO story”
Listen at 11:14
AI companies face falling revenue per token alongside potentially rising costs and depreciation.
“the revenue per token is coming down, but potentially the cost per token. And the total depreciation that comes through is increasing”
Listen at 12:11
NVIDIA’s 2027 demand and revenue growth are largely locked in.
“2027 is most likely locked in for NVIDIA and put across the whole supply chain”
Listen at 13:22
Chinese AI companies remain behind OpenAI and Anthropic at the frontier.
“they're not really at the frontier the way we see with OpenAI and Anthropic”
Listen at 23:55
SoftBank shares fell after OpenAI delayed its IPO.
“SoftBank shares got hit”
Listen at 25:03
SK Hynix fell about 7% and Samsung fell about 5% amid AI spending concerns.
“SK Hynix, which has been the leader in HPM, was down about 7%. Samsung is down about 5% too”
Listen at 26:10
Chinese concerns include AI chatbots manipulating emotions or political views.
“the AI chatbots could emotionally manipulate people or change their political views”
Listen at 31:11
Donald Trump appears opposed to regulation that would slow AI development.
“he has no interest, it seems, in any kind of regulation that is going to slow the pace of development”
Listen at 31:30
Statements are attributed to the speaker as said on the episode and reflect their view at the time, not PodLume's. They are not advice.