Oct 24, 2026 · 1h 33m
OpenAI’s valuation tests AI’s growth story
OpenAI’s $1.25T Raise, Starlink Mobile & Zuck’s Tax Deal | EP 2354
The episode connects a proposed $1.25 trillion OpenAI valuation to competitive pressure, fragile safety oversight, and the broader consequences of concentrated technology power.
- 1OpenAI’s reported $30 billion raise forces investors to test extraordinary growth assumptions against revenue multiples and competitive threats.
- 2Starlink Mobile could intensify wireless competition by combining satellite connectivity, dense cell infrastructure, and falling consumer prices.
- 3Paid AI safety audits raise a structural question: can evaluators remain independent when the labs they assess shape the field?
Don't miss
The hosts connect Meter’s findings and its founder’s prior AI-lab work to a broader argument for structurally independent safety oversight.
The brief
Jason and Lon begin with founder training and AI-assisted editorial work, arguing that useful automation still depends on source material, style discipline, and human review.
OpenAI’s reported $30 billion raise at a $1.25 trillion valuation becomes a test of growth assumptions, revenue quality, liquidation preferences, and competitive threats from other AI platforms.
The hosts then examine SpaceX’s reported Starlink Mobile ambitions, asking whether satellite links, cell sites, mesh networks, and lower prices could reshape U.S. wireless competition.
Findings from Meter and Redwood Research prompt a sharper concern: paid external audits may not be independent enough when AI labs and safety researchers share such close ties.
The conversation widens to AI liability, California’s proposed billionaire tax, Meta’s reported arrangement for Mark Zuckerberg, and the difference between startup hype and criminal fraud.
Off-duty recommendations turn to A Short Stay in Hell, The Library of Babel, and The Eureka Machine as lenses on ambition, discovery, and impossible scale.
What was said on this episode
23 statements · 15 positive · 7 negative · 1 neutral
OpenAI’s reported valuation equals roughly 30 times annualized revenue.
“It's 30 times price to sales ratio. Price. valuation to sales, sales top line revenue, not profits, not gross profit, not EBITDA”
Listen at 24:23
OpenAI’s reported valuation appears appropriately priced relative to revenue.
“I think they've right priced it”
Listen at 25:04
OpenAI could become a good ten-year investment if revenue keeps doubling or tripling annually.
“If in the year of our Lord, October 23rd, 2026, it's at 30 times price of sales and the thing's doubling every year, tripling every year, I think it closes the gap and it's probably a good buy if you were holding it for 10 years.”
Listen at 25:51
OpenAI deserves one of two concentrated positions among leading AI companies.
“I think Open Eye gets one of the slots.”
Listen at 29:14
Anthropic currently ships a slightly better product than OpenAI.
“Anthropic ships better product, I think, slightly better.”
Listen at 29:44
Young investors with long horizons should allocate 30–40% to OpenAI, SpaceX, and Anthropic.
“Clearly, put 30%, 40% into those three names if you could.”
Listen at 31:48
Starlink Mobile’s total addressable market may exceed current expectations and increase competition.
“The TAM here might be a little bit bigger than people suspect and that this is incredible for competition.”
Listen at 35:17
Starlink Mobile competition will reduce wireless and connectivity prices.
“I think it's going to... cause prices to go down.”
Listen at 38:53
Amazon Leo could become available to consumers around 2028.
“You reasonably could get this in 2028.”
Listen at 42:36
Paid auditors may allow companies time to resolve issues before including them in final reports.
“what they will say is like, hey, we're going to do this contract. We want 90 days to or nine days or nine hours to resolve anything.”
Listen at 48:36
Cybersecurity is the most important near-term AI challenge.
“The most important thing is cyber.”
Listen at 54:13
AI providers should conduct stronger customer due diligence.
“They need to know their customer more often.”
Listen at 54:29
Some earlier AI interactions encouraged or led users toward tragic outcomes.
“But in some of these earlier interactions, mistakes were definitely made. Settlements will be made. And it was kind of encouraging or leading people toward these tragic outcomes.”
Listen at 57:17
AI safety audits should be conducted by independent law-enforcement and intelligence experts.
“They need to hire FBI agents who worked in cybercrime, CIA, and they need to have them go do these audits.”
Listen at 1:02:43
AI oversight needs an independent public-sector talent pipeline outside frontier labs.
“There should just be an independent... Right now, if you're an AI genius and you graduate Stanford, you just get sucked up immediately by one of these labs.”
Listen at 1:03:08
California’s proposed billionaire tax would function as an asset seizure.
“It's an asset seizure.”
Listen at 1:03:47
Very high tax rates can compel affluent residents to leave a jurisdiction.
“When taxes go above 40%, 50%, you really do need to think as an affluent person, what am I getting for these dollars? And certainly when it hits 60%, like Denmark, you basically have to leave. You have no choice but to leave.”
Listen at 1:06:16
Taxing margin loans could mitigate billionaire-tax concerns over illiquid shares.
“This whole thing would be mitigated if they just said, anytime you take out a margin loan, you pay the taxes on that margin loan or some percentage of tax as if you had sold your shares.”
Listen at 1:07:06
Jason predicts California’s billionaire tax will pass despite prediction-market odds.
“I am- 90% certain it would. It will.”
Listen at 1:07:49
The Zuckerberg-Meta arrangement will be used politically to promote the billionaire tax.
“This is going to be weaponized to get the billionaire tax to pass.”
Listen at 1:12:57
The federal government should raise the minimum wage to $15.
“raise the minimum wage to 15”
Listen at 1:16:09
The federal government should ensure construction of one million homes annually.
“you're going to build 1 million new homes per year in the United States”
Listen at 1:16:14
Claiming an ongoing partnership without one would be materially false in an investment context.
“Materially false would be, and we've got a product coming out Q3 next year. And it'd be like, well, you only talked to that person one time, you never followed up with them, and that person doesn't have a contract from you or anything. And if you said, we have a partnership with them, that would be materially false if there wasn't an ongoing partnership.”
Listen at 1:27:16
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.
Books & mentions
Some links are affiliate links — PodLume may earn a commission if you buy.
Weekly email
State of the Pod, every Monday
The week's most-discussed names and stories from podcasts, with the moments behind them. Free, five minutes, unsubscribe in one click.