The show was taped the day before SpaceX's listing (June 11). The backdrop is chaotic: the US has begun airstrikes on Iranian targets, Brent crude is approaching $100, and the S&P hit a five-week low, all in the same week SpaceX is to complete the largest IPO in history.

On SpaceX, D.A. Davidson's head of technology research Gil Luria offers an interesting framework: he has tracked Musk-family companies for over a decade and found their valuations are never tied to fundamentals, the secret being a 'three-layer narrative' — the business in hand, the business on the way, and the big-dream business. Tesla is selling cars → Robotaxi → Optimus robots; SpaceX is Starlink → orbital data centers → sending a million people to Mars (the S-1 even literally says 'extending consciousness to the stars'). As long as there's always a bigger story behind it, the valuation can be 95x price-to-sales, or 50x, or 150x — 'it all depends on how many people buy the dream.' He cautions that beyond the day-one euphoria there are further milestones like large-scale lockup expirations and index inclusion, and this stock's volatility will exceed that of any company of comparable size in history; it's already draining liquidity from the whole market — being oversubscribed 4x means investors have $300 billion of cash ready, money pulled out of other AI stocks.

The bigger picture is a 'scramble for capital': OpenAI filed confidentially for an IPO in a hurry right after Anthropic (the official line is 'we'd rather stay in the private market, but understand we may need to go public,' which Gil translates as: afraid of Anthropic pre-empting the disclosure terms and valuation narrative); Google issued $85 billion in stock this week, Oracle announced a $20 billion offering plus $20 billion of debt (its stock fell 8% after earnings), and Meta is rumored to be considering similar moves. Ed's read leans bearish: everyone rushing to sell stock now suggests they think they won't get this price later — 'which makes me think we may be right at the top.' Gil is relatively more optimistic: the models really are getting stronger (he mentions user feedback that Anthropic's just-released Fable is another big leap forward), and as long as consumers keep paying subscriptions and enterprises keep buying tokens, the growth is still there; scrambling for capital isn't being bearish on yourself, but 'this is a game of kings and queens — you can't sit at the table without tens or hundreds of billions of dollars.'

On inflation, Moody's chief economist Mark Zandi analyzes May's 4.2% CPI (the hottest in three years): the bulk is the Iran war pushing up energy, but tariff pass-through, service-price stickiness from tighter immigration, and AI's impact are all in there too — AI is fiercely pulling demand (data-center construction, wealth-effect spending, electricity prices, chip prices) but hasn't yet delivered supply-side productivity gains, which is itself an inflationary factor. The market has already priced the odds of a rate hike this year at 50%. The base case is that Trump reaches a deal with Iran and oil prices retreat, but 'almost every other scenario is worse': global crude inventories are already as low as the early 1980s, and if the Strait stays blocked there will be physical shortages and unanchored inflation expectations — that's the stagflationary-recession script. On Trump's line 'I love inflation,' Zandi bluntly says 'there's nothing worse than inflation — it's lose-lose-lose.'

On Apple, The Verge alum Alex Heath brings observations from WWDC: the new Siri is driven by Google's Gemini via distillation + post-training (not the native Gemini experience), and it's indeed better than the old Siri, but 'the old Siri's bar was so low it was only fit for setting alarms'; there's no cross-conversation memory, no firm release date, and no launch in the EU or China, and the stock turned from an intraday record high to a decline, now down over 7%. He thinks Apple's real strategy is to make Siri a 'shell/harness' whose underlying model can be swapped at any time (Google today, maybe Claude or an open-source model tomorrow), and to accelerate once the technology matures — this is Apple's usual 'enter a cycle or two late' playbook; the real decade-long risks are agents making the operating system itself obsolete, and new hardware forms (OpenAI x Jony Ive, Meta glasses) impacting its services business.

In closing Ed makes his solo call: don't buy the SpaceX IPO. His prediction: the stock will surge about 25% on day one, because the investment banks can't let it fail (otherwise the OpenAI and Anthropic IPOs would all be called into question); but the valuation makes no sense, the insiders all know it, and once the lockup expires it will be one of the largest sell-off events in market history, cut in half within six months. 'Don't think you can win this game — by the time you buy, the insiders are already selling.'

(For comparative reading: the previous piece, Cathie Wood's take, is the bullish view on the same event; the two together form the full bull-bear panorama of the SpaceX IPO.)