Oct 14, 2026 · 38 min
Billionaire cash meets AI’s profitability test
Elon Musk Spent $366M On The Midterms — Will It Work?
The episode connects the diminishing power of political spending with the financial risks emerging across the AI infrastructure boom.
- 1Bradley Tusk argues that billionaire campaign spending buys access but loses electoral force as voters become saturated.
- 2Gil Luria separates durable semiconductor demand from speculative AI infrastructure financed on fragile economics.
- 3The episode questions whether AI companies’ reported margins reflect the full cost of training and building models.
Don't miss
Gil Luria warns that a downturn in the AI cycle could leave heavily indebted CoreWeave’s equity worthless.
The brief
Bradley Tusk argues that Elon Musk’s $366 million midterm spending can buy access and influence, but repeated advertising eventually overwhelms voters and weakens each additional dollar.
The political discussion turns to backlash, divided government, and a system that keeps attracting billionaire money even as spending becomes less effective.
Gil Luria examines the semiconductor sell-off and asks whether AI revenues can support the ecosystem’s enormous costs, especially when companies separate inference from training expenses.
Luria draws a sharp line between established chip demand and speculative infrastructure, warning that heavily indebted CoreWeave could be devastated if the AI cycle turns.
The episode closes with a report that Kimberly Guilfoyle was asked to resign as ambassador to Greece amid allegations involving donor access and lobbying promises.
What was said on this episode
25 statements · 12 positive · 7 negative · 3 mixed · 3 neutral
Campaign money matters less than commonly perceived.
“I think that the role, the perception of the import of money in campaigns is overrated.”
Listen at 4:24
Recent candidates won through energy and cultural fit, not spending.
“they won because they had the most energy and because they fit the zeitgeist of the moment.”
Listen at 5:45
Most of Musk’s midterm spending will be wasted.
“I think it largely will be for waste.”
Listen at 7:10
Musk’s spending will not prevent Republicans losing the House.
“his money is not going to change the fact that the House is going to flip.”
Listen at 7:48
Early campaign spending helps, but additional spending quickly has diminishing impact.
“The first money in might actually be impactful, but then it becomes less impactful.”
Listen at 9:02
Targeted six-million-dollar spending can help defeat candidates.
“money can be helpful. But that whole effort so far has been... Six million dollars.”
Listen at 10:24
Social media is more electorally effective than expensive broadcast television.
“makes things that are free or very cheap, like social media, really impactful and makes things that are highly expensive and outdated, like broadcast television, much less impactful.”
Listen at 10:46
Democrats will win approximately 52 Senate seats in the midterms.
“roughly 52 Democrats will end up with about 52 seats.”
Listen at 11:07
The midterms will have little impact on the technology sector.
“there's actually not going to be much impact on tech in the midterms”
Listen at 11:38
Technology policy outcomes will be similar regardless of the midterm winner.
“you're going to see relatively uniform results that would have been exactly the same kind of regardless of who wins or loses in a few weeks.”
Listen at 12:06
Sam Altman’s opposition unintentionally helped Alex Boris’s campaign.
“Sam Altman inadvertently in trying to kill Boris ended up helping them.”
Listen at 12:57
Opposition from unpopular technology companies can backfire electorally.
“it could actually have a collateral damage effect.”
Listen at 13:14
Political spending will rise while becoming less effective.
“The amount might continue to increase. I think the efficacy will continue to decrease.”
Listen at 13:21
AI value-chain companies report strong results and improving market conditions.
“all we've heard over the last several weeks, I want to just be clear about that, from all the companies throughout the AI value chain, is not only are they reporting very good results, they're saying the results and the market is better than it was three months ago.”
Listen at 19:41
Anthropic and OpenAI revenue run rates will exceed $100 billion combined.
“combining their revenue run rate with OpenAI's revenue run rate, we're going to be well in excess of $100 billion.”
Listen at 21:56
AI model revenue growth and profitability will likely continue over the next year.
“I think over the next 12 months, that's the likely trend.”
Listen at 23:25
Anthropic and OpenAI may never become fully profitable overall.
“it is completely speculative. whether they will ever become completely profitable on an overall basis”
Listen at 27:12
OpenAI and Anthropic will likely retain the high-end AI model market.
“it's likely that they will have the high end of the market.”
Listen at 29:52
Frontier AI models will likely serve the most mission-critical systems.
“The most potent tokens, the ones used for the most mission-critical systems are still likely to come from the frontier, which these two companies dominate.”
Listen at 29:56
If the AI cycle lasts through 2030, four major chip companies will capture most profits.
“if the cycle continues to 2030, those four companies will grab most of the profit.”
Listen at 31:35
The best semiconductor opportunities are NVIDIA, Broadcom, SK Hynix, and Micron.
“realistically speaking, that's where the opportunities are, are with those four companies”
Listen at 31:51
A downturn in the AI cycle could reduce CoreWeave’s equity value to zero.
“if the AI cycle sputters or certainly if it doesn't or if it turns over, CoreWeave is going to be the worst, the first to go. The equity value is going to be zero.”
Listen at 32:41
CoreWeave may fail to make interest payments and be wiped out.
“If things don't go well, they will not be able to make their interest payments. They're going to get wiped out.”
Listen at 33:31
Luria favors Micron, NVIDIA, Broadcom, and SK Hynix among semiconductor stocks.
“it's primarily Micron, NVIDIA, to some extent Broadcom and SK Hynix.”
Listen at 33:49
Several software companies are benefiting from improving AI tools.
“You've got Palantir, Snowflake, Datadog, Shopify, Brownstrike, extraordinarily good companies that are all benefiting from AI tools getting better.”
Listen at 34:00
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.