Big Tech hides the scale of its leveraged AI bet

Big Tech’s Insane Hidden AI Spending, Ranking Anthropic vs. OpenAI, AI For Travel Debate

The episode connects opaque infrastructure financing, uncertain AI demand, and a widening competitive gap between Anthropic and OpenAI.

3 key takeaways

1External financing and leases can make Big Tech’s AI infrastructure commitments look smaller than their underlying obligations.
2Anthropic appears to be gaining momentum while OpenAI faces slower growth, deeper losses, and senior executive departures.
3Complex travel planning may test AI agents more meaningfully than standardized benchmarks because it exposes failures in real-world logistics.

Don't miss

The hosts argue that complicated international travel could be a more revealing everyday evaluation of AI agents than conventional benchmarks.

The brief

Alex Kantrowitz and returning guest Ranjan Roy begin with a question beneath the AI boom: how much infrastructure spending is Big Tech keeping off its balance sheet?

External investors, leases, guarantees, and financing vehicles can shift data-center obligations away from corporate balance sheets, raising questions about who ultimately bears the risk.

The hosts weigh whether AI demand can repay this leveraged buildout, or whether many companies are making the same enormous call option on AGI at once.

Anthropic appears to be pulling ahead as its reported revenue momentum contrasts with OpenAI’s slower growth, deepening losses, executive turnover, and unsettled strategy.

The conversation ends with a practical test for AI agents: complicated international travel, where coordination and execution may matter more than benchmark scores.

What was said on this episode

17 statements · 6 positive · 8 negative · 3 mixed

Ranjan Royon Big Tech AI infrastructure financingNegative5:41
External financing shifts AI infrastructure investment risk away from Big Tech balance sheets.
“So it's a massive investment that you've managed to creatively push off your balance sheet and move and spread that risk to other people or other pools of capital.”
Listen at 5:41
Ranjan Royon Big Tech off-balance-sheet financingMixed19:17
Obscuring AI obligations is tactically logical for Big Tech companies.
“So from every purely tactical level, this is logical. unfortunately, but it is.”
Listen at 19:17
Ranjan Royon Big Tech AI investmentNegative25:07
Major technology companies are making the same leveraged AI investment bet.
“The danger is very clear that these are the largest companies in the world and they're taking a levered bet. And they're all taking the same levered bet.”
Listen at 25:07
Ranjan Royon AI demand and agentic AIPositive25:33
AI demand will surge and agentic AI will reshape the economy in coming years.
“I firmly believe that AI demand will explode and the economy is going to get agentified and all of that's going to happen in the years to come.”
Listen at 25:33
Alex Kantrowitzon AI infrastructure financingNegative29:57
The current AI infrastructure financing trajectory is unfavorable.
“The substance is that this is not going in a good direction.”
Listen at 29:57
Ranjan Royon AI infrastructure financing cycleNegative30:15
The AI infrastructure financing cycle will not unfold smoothly.
“I don't think this plays out completely smoothly.”
Listen at 30:15
Ranjan Royon Anthropic and OpenAI IPOsNegative31:44
OpenAI and Anthropic IPOs could disappoint within six to twelve months.
“six to 12 months, OpenAI and Anthropics IPOs are duds.”
Listen at 31:44
Alex Kantrowitzon Artificial IntelligenceMixed33:58
AI has three principal outcomes: beneficial success, harmful success, or failure.
“I'm starting to think there's really only three true outcomes for AI.”
Listen at 33:58
Ranjan Royon Anthropic IPONegative37:29
AI’s medical potential does not establish that Anthropic’s IPO should succeed.
“I don't think that tells a story about why Anthropics IPO should succeed or you should like Anthropics specifically.”
Listen at 37:29
Ranjan Royon Anthropic revenue figuresNegative44:38
Anthropic’s reported revenue figures remain insufficiently reliable for evaluation.
“I'm not comfortable with these numbers yet.”
Listen at 44:38
Alex Kantrowitzon OpenAI CodexNegative50:12
OpenAI is behind Anthropic in coding and is refining Codex after an initial release.
“they got their first draft out of it before, you know, sometime in July. And now they're going to refine.”
Listen at 50:12
Alex Kantrowitzon OpenAI and Anthropic technologyMixed55:25
OpenAI and Anthropic’s technologies and offerings remain immature but rapidly expanding.
“The technology that they're developing is not mature. It's still growing. The capabilities are growing fast. The offering is growing fast.”
Listen at 55:25
Alex Kantrowitzon OpenAI and AnthropicPositive57:17
Anthropic currently appears to be widening its lead over OpenAI.
“at least when you look at the state of play right now, yeah, Anthropic is putting some distance between itself and OpenAI.”
Listen at 57:17
Alex Kantrowitzon Travel as an AI evaluationPositive58:23
Travel is an excellent real-world test of AI capabilities.
“travel when it comes to assessing AI's capabilities is actually an excellent proving ground.”
Listen at 58:23
Ranjan Royon Travel as an AI evaluationPositive1:01:16
Travel is a valuable evaluation framework for AI systems.
“I like the framing travel as the eval. I actually kind of love that, in fact.”
Listen at 1:01:16
Ranjan Royon AI travel planningPositive1:03:00
AI successfully converted travel requirements into a suitable destination recommendation.
“AI got me there. It was literally set of requirements.”
Listen at 1:03:00
Alex Kantrowitzon AI travel logisticsPositive1:04:00
AI made Alex’s complex international travel logistics seamless.
“it was just logistically so, so seamless.”
Listen at 1:04:00

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.