Oct 22, 2026 · 26 min

IBM warns AI spending needs clearer returns

Did IBM Just Tell the Truth About AI?

The episode connects semiconductor demand, infrastructure costs, and enterprise caution to a broader test of whether AI investment can keep accelerating.

3 key takeaways

1Lam Research benefits from record services revenue and advanced packaging demand, but semiconductor equipment remains exposed to cycles and China.
2IBM’s strong results meet a tougher enterprise spending environment, as customers demand clearer returns before expanding AI budgets.
3Helium, fertilizer, memory, and infrastructure costs could reshape the economics behind semiconductor manufacturing and on-premise AI deployment.

Don't miss

IBM’s results look solid, but the hosts focus on Arvind Krishna’s warning that customers need clearer returns before committing more AI budgets.

The brief

Lam Research’s record performance and recurring services revenue suggest durable semiconductor demand, but the hosts question how long the current cycle can last.

Advanced packaging—stacking and connecting multiple chips—has become increasingly important for AI processors and high-bandwidth memory, strengthening the equipment case.

The discussion widens from chip demand to supply risks: Gulf disruptions could lift helium and fertilizer costs, adding pressure across manufacturing and agriculture.

IBM delivers revenue growth and strong free cash flow, yet CEO Arvind Krishna’s warning points to a sharper constraint: enterprises want clearer returns from AI spending.

The final valuation debate splits the hosts: Matt Frankel sees hidden value and a quantum-computing call option, while Lou Whiteman remains cautious on technology stocks.

What was said on this episode

18 statements · 7 positive · 8 negative · 2 mixed · 1 neutral

Matt Frankelon Lam Research toolsPositive1:55
Each Lam Research tool generates years of subsequent service revenue.
“every tool that LAMM sells creates years of service revenue”
Listen at 1:55
Matt Frankelon Advanced packagingPositive4:22
Advanced packaging demand should increase as AI chips become more sophisticated.
“as AI chips evolve, get more sophisticated, things like this should only increase”
Listen at 4:22
Matt Frankelon Lam ResearchPositive7:00
Matt would modestly buy Lam Research after its decline.
“It's a dip that I would nibble on”
Listen at 7:00
Lou Whitemanon Lam ResearchPositive7:14
Lou does not think investors should sell Lam Research because of cyclicality yet.
“I don't think it's time to sell on the cycle”
Listen at 7:14
Lou Whitemanon Lam ResearchNeutral7:56
Lou prefers waiting for a better Lam Research entry price.
“I'm going to be stubborn and wait and see if I can't get a better deal”
Listen at 7:56
Matt Frankelon Helium in semiconductor manufacturingNegative11:14
Helium has no good substitute in semiconductor manufacturing.
“There's no good substitute for it”
Listen at 11:14
Lou Whitemanon 2027 farming conditionsNegative13:40
The Iran-related fertilizer and fuel pressures will make 2027 terrible for farming.
“It's almost baked in that 2027 is going to be a terrible year for farming”
Listen at 13:40
Lou Whitemanon U.S. companiesNegative14:22
Gulf-related agricultural pressures will eventually affect U.S. companies.
“the implications from all of this on the global Western economy are going to hit U.S. companies over time”
Listen at 14:22
Lou Whitemanon 2027 economyNegative14:49
Lou expects 2027 to be economically difficult.
“I think 2027 is going to be a slog”
Listen at 14:49
Matt Frankelon AI tradeMixed15:41
The next AI investment upswing will be selective and company-specific.
“I see the next leg up in the AI trade. as existing, but being a lot more selective and very company-specific”
Listen at 15:41
Lou Whitemanon Enterprise market pricingNegative18:32
Enterprise customers are pushing back more on pricing.
“there is more pushback on pricing in the enterprise market”
Listen at 18:32
Lou Whitemanon IBMPositive18:43
IBM is unusually positioned to provide an honest view of AI demand.
“IBM is uniquely positioned to tell the truth here”
Listen at 18:43
Matt Frankelon IBM software and consulting servicesNegative20:03
Demand for IBM’s lucrative software and consulting services is limited and selective.
“There's not unlimited demand. It has to be focused.”
Listen at 20:03
Matt Frankelon IBM fourth-quarter resultsNegative21:01
IBM’s unchanged guidance after a third-quarter beat implies a weak fourth quarter.
“When you beat expectations in a third quarter and keep your guidance flat, you're implying that the fourth quarter is going to be weak.”
Listen at 21:01
Matt Frankelon IBM strategyPositive21:38
Matt is more confident in IBM’s strategy over the next one or two years.
“I'm more confident in IBM's strategy over the next year or two than I was before the report.”
Listen at 21:38
Lou Whitemanon AI industry economicsMixed23:22
AI economics must eventually favor either infrastructure providers, hyperscalers, or enterprise efficiencies.
“Either the picks and shovels win, the hyperscalers win, or companies see greater efficiencies.”
Listen at 23:22
Matt Frankelon IBMPositive24:41
Matt considers IBM his best call option on quantum computing.
“I consider it my best. Call option on the quantum computing trend.”
Listen at 24:41
Lou Whitemanon Technology stocksNegative24:52
Lou currently avoids technology stocks.
“I'm not really interested in anything tech right now”
Listen at 24:52

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.

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.