TWiST News: Subscription Economics, Density's Waffle, and The Cost of AV Fleets | E2027

Episode Summary

Executive Summary: The episode centers on three big themes: a favorable FTC shift on subscription cancellations, Density’s launch of Waffle—a cheaper, self-installable room-occupancy sensor for offices and other spaces—and the growing infrastructure race around robo-taxis and nuclear power. The hosts argue that lowering friction and improving unit economics unlock adoption, and that data, energy, and physical-world intelligence are becoming core competitive advantages.

Main Topics: FTC subscription rules and “click to cancel” (Priority: 5/5): Jason and Alex discuss new FTC rules that make it easier for consumers to cancel subscriptions. Jason is unusually supportive, arguing it reduces dark patterns and improves trust, while acknowledging the tension with subscription businesses that rely on inertia. Density’s Waffle product launch (Priority: 5/5): Andrew Farah introduces Waffle, a self-installable, lower-cost occupancy sensor that expands Density’s room-level sensing into more deployments. The product is positioned as a major step in reducing installation cost and broadening use cases beyond enterprise offices. Physical-space intelligence and workplace optimization (Priority: 4/5): The conversation explores how occupancy data can optimize conference rooms, seating, cleaning, HVAC, wayfinding, and real estate decisions. The hosts emphasize that most buildings are still effectively unmeasured and that better sensing can change how space is planned and used. Robo-taxi fleet modeling and AV economics (Priority: 4/5): Alex previews a spreadsheet model estimating how many autonomous vehicles would be needed to replace ride-hailing and broader transportation demand. The discussion focuses on trip frequency, production scale, costs, and how efficiency per vehicle drives fleet economics. Nuclear power resurgence and hyperscaler demand (Priority: 5/5): The hosts discuss Google’s and Amazon’s nuclear energy commitments as proof that large tech companies are creating a real market for small modular reactors (SMRs). They argue that AI, data centers, and grid needs have made nuclear politically and economically viable again. Solar, batteries, and resilience (Priority: 3/5): The episode tees up a follow-on discussion about solar oversupply, storage, and off-grid resilience. Jason and Alex frame distributed energy as both a globalization issue and a personal independence issue, especially in the face of weather, grid, and supply-chain risks.

Key Arguments: Easier cancellation rules are good for consumers and can also improve businesses by surfacing real signals about product fit and churn. Subscription businesses should compete on value, not by trapping users with dark patterns or hidden cancellation friction. Office space is still massively under-measured; occupancy intelligence can materially reduce real-estate, cleaning, and energy waste. Waffle lowers deployment friction by removing the biggest cost barrier: installation, which should accelerate adoption. Privacy-preserving sensing is preferable to cameras because it enables adoption without making people feel surveilled. Autonomous fleets will only scale meaningfully if vehicle utilization and production economics improve dramatically. Nuclear is back because data centers and AI need reliable baseload power, and hyperscalers are now direct customers for new reactors. Solar and storage are becoming cheap enough to reshape resilience, grid planning, and household energy independence. Redundancy and resilience—in internet, power, and food systems—are becoming more valuable to businesses and households.

Data Points: One-time office sensor deployment cost: About $1,000 all-in per room - Jason and Andrew discuss the previous Density deployment economics, including hardware, software, and installation. Installation share of deployment cost: 65% - Andrew says most of the old system cost came from installation rather than the technology itself. Waffle pricing: $150 per unit and $8/month - Jason estimates the new self-install product’s price during the product demo. Real-time latency: Sub-700 milliseconds - Andrew says Waffle turns occupancy status red almost instantly when someone enters a room. Meeting room usage by one person: 53% - Andrew cites a figure showing over half of meeting-room use was by a single person over the last year. Office space measured globally: About 3 trillion square feet - Andrew describes the scale of physical space Density aims to measure. Unmeasured space: About 99.10% unmeasured - Andrew says nearly all physical space lacks usage data. Carbon emissions from buildings: 39% of global carbon emissions - Andrew connects occupancy sensing to sustainability and energy optimization. Net new buildings per day: 13,000 - Andrew cites daily global building additions when discussing the climate opportunity. U.S. annual trips: 400 billion - Used in the robo-taxi fleet model based on the National Household Travel Survey 2022. Uber and Lyft trips in the U.S.: 4 billion - Alex uses this as a smaller subset in the AV model. Public transit trips in the U.S.: 20 billion - Alex includes public transit demand in the robo-taxi modeling spreadsheet. Trips per day per vehicle: 20 - The model assumes roughly 20 rides per day for a robo-taxi after downtime, cleaning, and charging. Maintenance downtime assumption: 90 days per year - Alex models 6 hours per day of downtime, leaving 275 service days. Waymo vehicle cost estimate: $75,000 to $120,000 per car - The hosts discuss high-cost autonomous vehicle assumptions in the fleet model. Lower-cost AV estimate: $30,000 per vehicle - Jason references Tesla/BYD/Baidu-style cost assumptions for the model. U.S. fleet replacement cost: $2.2 trillion to $9 trillion - Estimated cost range to replace major vehicle demand with autonomous fleets. Google nuclear deal: First corporate agreement to purchase power from multiple SMRs - The hosts describe Google’s deal with Kairos as a landmark for nuclear energy.

Pivotal Quotes: "What did it cost? No exit market for four years." — Jason Calacanis: Jason jokes that FTC cancellation reforms are worth a less painful startup exit environment. "We don’t know how any of it is used. Something like 99.10% of it is unmeasured." — Andrew Farah: Andrew describes the scale of the occupancy-data opportunity for physical space. "If we could keep the cost curve on the bottom end, you can rapidly accelerate the adoption of new infrastructure." — Andrew Farah: Andrew explains why lowering deployment cost is essential for market expansion.

Implications: The episode argues that friction reduction—whether in subscriptions, office sensing, robo-taxi production, or energy infrastructure—drives adoption. Listeners should expect more privacy-safe sensing, AI-era power demand, and data-driven physical infrastructure investment.