Nov 3, 2026 · 29 min

AI economics collide with surveillance and orbital manufacturing

Apple & Palantir Earnings, Elon Caps Grok Bot Usage, SpaceX’s Starfall | Diet TBPN

The episode connects AI’s rising infrastructure costs and security risks with a broader race to build new computing and manufacturing systems.

3 key takeaways

1Earnings strength is colliding with fears that future margins will compress across major technology companies.
2AI search, agent access, and surveillance are expanding faster than the systems designed to control their risks.
3SpaceX’s Starfall points toward routine microgravity access for biomedical, pharmaceutical, and materials research.

Don't miss

Jordi’s location-identification exercise, using clues from airplane audio and Doppler signatures, turns an abstract surveillance risk into a concrete one.

The brief

John Coogan and Jordi Hays open remotely, then turn from Apple and Palantir earnings to the larger question of whether strong results can outrun margin pressure.

Google’s AI search strategy sits at the center of a frontier-model race, while enterprise spending increasingly favors OpenAI and Anthropic and Gemini becomes embedded in Search.

Elon Musk’s warning about GrokBot’s serving costs leads into a broader discussion of AI economics, including why usage limits may matter even for a wealthy company.

An exercise identifying Jordi’s location from airplane audio becomes a warning about AI-enabled surveillance: long recordings can disclose more than people realize.

CrowdStrike’s AI-security acquisitions and OpenAI’s partnership with Cathedral Systems frame agent security as urgent once software can access sensitive credentials and systems.

The episode closes on SpaceX’s Starfall plans for routine microgravity access, linking orbital manufacturing to biomedical research, pharmaceuticals, and materials science.

What was said on this episode

16 statements · 7 positive · 7 negative · 1 mixed · 1 neutral

John Cooganon AI ModePositive4:40
Google AI Mode is increasingly becoming the default search experience.
“AI mode is increasingly the default Google experience.”
Listen at 4:40
John Cooganon Google business modelMixed4:58
Google will monetize AI search and shift toward cloud and chip businesses.
“Google will milk that for as much revenue as they possibly can and transition from a search company to a cloud service and chip company as traffic from humans. It becomes more scarce.”
Listen at 4:58
John Cooganon Google AI searchNegative5:12
Google AI search creates a major headwind for non-destination media companies.
“This is a massive headwind to all media-type companies who are not destinations.”
Listen at 5:12
Jordi Hayson GooglePositive7:08
Google is difficult to view bearishly if AI becomes large and important.
“if you think AI is real and going to be big, it's very, very hard to make a super coherent bear case”
Listen at 7:08
John Cooganon Anthropic and OpenAI enterprise LLM API spendingPositive7:50
Anthropic and OpenAI’s share of enterprise LLM API spending rose to 86%.
“Anthropic and OpenAI now account for 86 percent of enterprise LLM API spend up from 67 percent.”
Listen at 7:50
John Cooganon Google enterprise LLM market shareNegative8:00
Google’s enterprise LLM-builder market share fell by more than 50%.
“Google's market share went down more than 50 percent in terms of not in terms of LLM usage for consumer, but but for builders in the enterprise.”
Listen at 8:00
Jordi Hayson GooglePositive8:35
Google remains one of the greatest businesses in history.
“this is still one of the greatest businesses in history.”
Listen at 8:35
Jordi Hayson GooglePositive9:46
Google has substantial upside if it remains competitive in cloud and chips.
“as long as they can stay competitive around the cloud and chip business, you know, there's a lot of upside.”
Listen at 9:46
John Cooganon Grok enterprise go-to-marketNegative14:28
Moving Grok from X entertainment into Slack enterprise use was a strategic stretch.
“The idea of like bridging that bridging from that, like the X, the everything app that's like mostly for fun and mostly for like gossip to like it's in my Slack now was always a little bit of a stretch”
Listen at 14:28
John Cooganon AI-enabled audio surveillanceNegative16:32
Audio-based location inference is an example of AI-enabled surveillance.
“this is a crazy example of like AI enabled surveillance.”
Listen at 16:32
John Cooganon Long-form audio recordingsNegative16:55
Long-form outdoor recordings can reveal a person’s location.
“this definitely could dox your location”
Listen at 16:55
John Cooganon X app interface designNeutral18:25
X uses interface patterns to increase user retention.
“there's so many different patterns to get the uh the the retention up for sure they're they're working on it”
Listen at 18:25
Jordi Hayson Rowdy snacksPositive19:03
Rowdy’s savory snack product will become a hit.
“it's going to be a hit.”
Listen at 19:03
John Cooganon Orbital manufacturing market demandPositive26:44
SpaceX’s entry indicates serious demand for orbital manufacturing services.
“there would be real demand behind it because SpaceX wouldn't be going after it unless there was serious demand.”
Listen at 26:44
John Cooganon AI agents and user credentialsNegative27:46
Users are giving AI agents access to their passwords and sensitive accounts.
“Everyone's giving Muse all their passwords, BOT all their passwords, ChatGPT all their passwords.”
Listen at 27:46
John Cooganon AI agent cybersecurityNegative27:51
Compromising credential-holding AI agents could expose email, services, and bank accounts.
“if you break into that, you're kind of in the email, you're in all the services, you're in the bank accounts pretty quickly.”
Listen at 27:51

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.