*Cash App is a financial services platform, not a bank. Banking services provided by Cash App’s bank partner(s). Prepaid debit cards issued by Sutton Bank, Member FDIC. Cash App Visa® Debit Flex Cards issued by Sutton Bank, Member FDIC, and The Bancorp Bank, N.A., pursuant to a license from Visa U.S.A. Inc. See terms and conditions for the Sutton prepaid card, Sutton debit flex card, and Bancorp debit flex card. Cash App Green features, Savings, Direct deposit, Round ups, Overdraft coverage and Discounts provided by Cash App, a Block, Inc. brand. Visit cash.app/legal/podcast for full disclosure.

Ketone IQ: Visit https://ketone.com/IMPACT for 30% OFF your subscription order.

ATT Business: Switch to AT&T Business at https://business.att.com

Incogni: Take your personal data back with Incogni! Use code IMPACT at the link below and get 60% off an annual plan: https://incogni.com/impact

Quo: Try for free PLUS get 20% off your first 6 months at https://quo.com/impact

Pipedrive: Get more leads and grow your business. Go to https://www.pipedrive.com/impact and get started with a 30-day free trial.

Netsuite: For the first time ever you can try NetSuite Next for free. If your revenues are at least in the seven figures, go to https://NetSuite.ai/Theory.

Butcherbox: Go to https://ButcherBox.com/IMPACT to get $20 off your first box, plus your choice of free ribeye, new york strip, or filet mignon in every box for a year — with free shipping always

The team breaks down a real and fast-moving shift in the AI industry: open-weight models, which were a rounding error (~7% of traffic) back in December, have exploded—accounting for 56% of tokens generated by August, with Vercel CEO Guillermo Rauch reporting a single-day record of 78.4% on September 19th, and spending on three Chinese labs (Moonshot, DeepSeek, z.ai) on that day topping spend on OpenAI. The host walks the money: open-weight share of dollars on the Vercel gateway went from under 4% in June to 8.6% in July to 14% in August, average price per token fell 23% in August alone, and developers are splitting the workload—hard, high-stakes work still goes to Anthropic (which has held 61%+ of gateway spend every month since December), everything else goes to whatever's cheapest. His read on what's breaking: not the frontier, but the low end of the big players—OpenAI's and Google's cheap "gateway" models are getting obliterated as people conclude "you're not as good as Anthropic, so I'll either pay for the best or use the cheap stuff." The real winners, he argues, are the chipmakers and cloud/inference providers (like Vercel), because value is shifting from owning the model to just providing access to the compute—the commoditization of intelligence. He frames the optimistic upside: companies will increasingly host their own open-weight models to keep their work proprietary and build a real edge, rather than everyone using the identical public tool. Ryan pushes the counter-thesis—that if OpenAI had stayed the 2015 nonprofit and gone all-in on open source (his analogy: how OBS open-sourced screen recording and dominated), the US might be further ahead of China. Tom steelmans it but argues the fatal flaw: someone has to build the "brains," the data centers are staggeringly expensive, and there's still no observed ceiling to how much smarter models get with more compute—so without the profit incentive to aggregate that capital, AI would have stalled out entirely. His synthesis lands close to the Musk-era idea: keep the frontier model paid to fund the enormous buildout, let older generations go open source to drive sub-frontier innovation, and don't let the government kill open weights. They also dig into China's logic—giving models away free as an authoritarian-enabled disruption play they can clamp down on later once they're manufacturing their own chips—and the cynical data angle that free Chinese models feed usage data straight back to labs that can close them off anytime. A sharp, first-principles look at where the AI value—and the risk—is actually moving.

The team steps back from any single flashpoint to map how the world's conflicts are increasingly linking into one interconnected crisis. The weekend's near-escalation—a Houthi missile barrage on Riyadh, an Aramco depot fire, a State Department warning to reconsider travel to the entire region, and Trump briefly in "deciding mode" before reportedly backing off—is treated as just one node in a much larger web. The host lays out the loosely-sourced but growing hypothesis tying it all together: a China-Russia-Iran-North Korea alignment whose logic is to keep the US bogged down in as many places as possible—so China can move on Taiwan (possibly accelerating a 2027 timeline if it feels cornered on AI and chips), Russia can test whether NATO's guarantees mean anything while a wartime economy gives Putin every incentive to keep the Ukraine stalemate going, and Europe openly tells its citizens to prepare for war, with Germany talking about the biggest military on the continent. Threaded through every conflict is the same economic engine: the disruption of the Strait of Hormuz and Bab el-Mandeb has pushed shipping costs from roughly $6–7 to $23+ per barrel, and that slow-building penalty cascades into fuel, food, and fertilizer crises—potentially a hard-winter scenario that pushes Europe to "escalate to de-escalate." The host's synthesis: calling it World War III is overstated, but these conflicts are genuinely connected and waiting on a spark. He traces the root cause to the erosion of the US as the "monolithic keeper of global peace"—a decline he pins on internal division and a watershed in Russia's Ukraine invasion—and argues America is now in a "Suez Canal moment": if it can't reopen Hormuz by force, the world's near-universal "back down when America says so" reflex collapses, and the US has to actually fight every skirmish at an unaffordable cost, on top of staggering debt. He frames it as a once-in-a-century shift in great-power politics unfolding at the exact moment AI emerges as the defining strategic technology, and closes on his recurring warning: regulating AI to death would hand China the opening to corner "intelligence itself." (A chat-prompted aside, which the host flags as a troubling hypothetical, entertains whether a large-enough war could be used to justify postponing an election—landing on Congress as the necessary check.)

The team breaks down Trump's declaration of an "ultimate deal" with Denmark over Greenland—and separates the victory lap from the substance. The host's read: much of what Trump is trumpeting was already in place under the 1951 US-Denmark agreement (which lets the US operate military bases there), and he's taking a modest update and blowing it up for low-information voters who only hear the headline. But there is a genuinely new and strategically significant element—Greenland can no longer allow foreign-adversary infrastructure, specifically shutting China out. He explains why that matters: China has been insinuating itself into the Arctic (icebreakers, a new shipping route, calling itself "Arctic-adjacent"), and Chinese-built infrastructure/tech has a documented pattern of quietly sending data home (he cites a Nordic test where an isolated Chinese car kept pinging China), so keeping Chinese equipment out of a territory that sits on the direct line of any Russian attack on the US—and any future "golden dome" early-warning/defense system—is a real national-security win, one he says Denmark, Greenland, the UK, and the US all seem to accept. His broader points: this is smart strategy people miss because they hate Trump, and he argues there's a real shift toward Americans who no longer believe the US has moral standing to be the strongest power—a view he rejects, favoring a US that reestablishes its moral core but refocuses on hemispheric defense rather than policing the world. His sharpest criticism is Trump's method: publicly humiliating counterparts ("be the little bro, pat them on the head") breeds the exact resentment now visible with Canada and Europe, when the same wins could be gotten privately while letting the other side save face. On the forward look, he frames Greenland's real friction as a potential US-Europe proxy fight, notes that Europe won't fully hitch itself to China because China is aligned with Russia (a "suicide mission" militarily), and addresses a viewer point that America is rapidly losing soft power—arguing the only durable way to bring allies back is genuine economic and military strength (people getting richer by aligning with America), not threats, which produce brief compliance and permanent option-seeking. His path: stabilize Iran, lean into the Abraham Accords and hemispheric energy (Venezuela + US), stop policing the world, calm the violent left-right whipsaw, and grow the economy through AI.

See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

The AI-powered Podcast Player

Save insights by tapping your headphones, chat with episodes, discover the best highlights - and more!

Get the app

Home Top podcasts Popular guests Top books Privacy Policy Terms of Use