By playing, you agree to uListen's Terms and Privacy Policy and YouTube's Terms of Service.

Can AI Fix Housing and Healthcare Affordability?

Housing and healthcare make up nearly half of household spending, yet both sectors are riddled with inefficiency and rising costs.

Minna SongguestErik TorenberghostAlex ImmermanguestTony Stoyanovguest

Aug 21, 202540mWatch on YouTube

Summaries, chapters, transcripts, and FAQs on this page are made by uListen, not YouTube.

CHAPTERS

Jump to publisher timestamps

Chapter summaries

Elise AI’s mission: autonomous buildings and cutting the biggest household costs
Minna lays out the core premise: the U.S. is millions of housing units short, and housing plus healthcare consume an enormous share of household spending. Elise AI’s north star is “fully autonomous buildings,” using AI to remove waste and friction in two of the economy’s most expensive sectors.
- •U.S. housing shortage framed at ~5 million units
- •Housing + healthcare cited as ~42% of typical household spend and ~40% of GDP
- •Goal: portfolios that can run core operations with minimal/no human intervention
- •Thesis: tech should reduce costs and improve service quality in essential sectors
Why a16z backed Elise AI: software finally can “eat” housing and healthcare
Erik and Alex explain the investment rationale: housing and healthcare have resisted the cost declines seen in tech-touched industries. They argue AI changes the equation by automating the operational, administrative, and communication burden—areas that historically blocked software adoption.
- •Historical pattern: tech lowers prices in most categories, but not housing/healthcare
- •AI enables a shift by automating high-volume admin and communications work
- •Elise AI started early (2017) and went deep in housing before expanding to healthcare
- •Customer delight + growth/efficiency metrics cited as validation
Affordability starts with supply—plus quick wins from better utilization
Minna anchors affordability primarily in supply: construction needs to rise materially to stop the shortage from worsening. In the near term, she argues operators can extract more value from existing stock by fixing broken demand capture—like failing to respond to inquiries—so units lease faster and stay occupied.
- •Supply is the biggest determinant of housing prices
- •Needed pace: ~1.8–2.0M units/year vs ~1.5M built last year; delivery must rise ~50%
- •Pipeline risk: projected decline in 2026+ deliveries
- •Operational inefficiency example: ~half of rental inquiries go unanswered
- •Elise AI data: buildings using their AI show ~2% higher occupancy than market
Why we’re building too little: zoning constraints and capital-return dynamics
The conversation turns to what blocks supply response: regulation (especially zoning) is part of it, but not sufficient. Minna emphasizes capital allocation—returns in housing must improve to attract more investment, and she argues operational efficiency can raise those returns and pull more capital into construction.
- •Zoning and regulatory constraints contribute to limited new supply
- •Even deregulation alone won’t fix shortages without more capital
- •Capital flows to higher returns; housing competes with other asset classes
- •Elise thesis: create “10x better operators” to increase profitability and attract capital
- •More capital into housing → more supply → affordability improvement
Will YIMBY reforms happen? Evidence from Minneapolis and supply response timing
Tony offers cautious optimism on regulatory reform, pointing to Minneapolis as a proof point where zoning changes correlated with faster supply growth and flatter rents. They discuss how quickly supply could respond if rules changed broadly, and the expectation that markets would build if allowed to.
- •Hopeful outlook: some cities are moving toward pro-building reforms
- •Minneapolis case: ending single-family zoning (2019) and stronger supply growth cited
- •Reported rent outcomes: Minneapolis rents flatter vs national increases over same period
- •If permitted, builders will build—full impact still takes years
- •More competition could spur innovation in building and operations
Where tech can move the needle now: labor as the biggest controllable cost
Minna explains why housing operations underperform on efficiency: the industry historically underinvested in technology, while labor and other costs (insurance, supply chain) have surged. With AI, the most controllable lever is labor—automating repetitive workflows and preventing downstream costs like legal issues or avoidable CapEx.
- •Real estate lagged tech investment, leaving large efficiency gaps
- •Labor described as the biggest controllable operating expense, worsened post-COVID
- •AI can automate inefficient administrative workflows to reduce staffing burden
- •Secondary savings: fewer legal/compliance issues; better preventative maintenance reduces CapEx
- •Tech offsets cost inflation even when supply is constrained
Efficiency tactics without new supply: faster turns, smaller units, and better connectivity
Responding to examples like SF and NYC, Tony outlines ways to increase affordability without adding units: improve utilization by turning and filling units faster, consider smaller apartments and shared amenities, and expand effective supply through infrastructure that connects surrounding regions.
- •SF vacancy (~3.5%) suggests room for utilization gains even without new units
- •Tech can reduce manual inefficiencies to cut time-to-lease and time-to-turn
- •Design ideas: smaller units, shared amenities, flexible layouts
- •Infrastructure links (e.g., Jersey–NYC) expand metro-area effective supply
- •These are partial fixes; long-term affordability still requires building more
The autonomous building vision: centralization and the new staffing frontier
Minna describes the end state: portfolios that run core operations autonomously, leaving humans primarily for physical work and mandated tasks. She gives examples of centralization—moving from on-site generalists to portfolio-level roles—where AI enables dramatic unit-per-employee leverage.
- •Goal: fully autonomous buildings/portfolios for core operations
- •Most on-site work is administrative and automatable; physical maintenance remains a limiter
- •Equity Residential example: reaching ~200 units per employee
- •Brookfield example: centralized model enabling one specialized role to support ~10,000 units
- •Achieving the frontier requires reimagining operator workflows, not just adding tools
What’s automated today: maintenance triage, leasing Q&A, self-guided tours, and paperwork
They enumerate concrete areas where AI is already delivering value: digitizing and optimizing maintenance workflows, automating leasing communications, enabling self-guided touring via smart access, and eliminating manual documentation. The reported outcomes include faster work-order completion and shorter lease-up cycles.
- •Maintenance ops: AI triage/prioritization/routing; completion times cut to <48 hours from 4–5 days
- •Leasing: AI handles repetitive questions and high-volume communications
- •Touring: smart locks/lockboxes plus AI engagement enables 24/7 touring and faster leasing
- •Time-to-lease improvement cited: ~30 days down to <14 days for some customers
- •Documentation automation reduces copy/paste administrative burden
Second-order automation and the human role: ecosystem orchestration + new careers
Tony introduces “ecosystem-level” optimization—sharing resources across buildings for bigger maintenance gains. Minna addresses workforce implications: roles don’t vanish so much as shift from menial admin to resident relationships, conflict resolution, specialization, and supervision of AI-driven systems; physical maintenance remains essential amid labor shortages.
- •Next wave: cross-building orchestration of people, parts, tools to improve utilization
- •AI can solve complex planning/scheduling across portfolios, not just per-building
- •Humans shift to community building, relationship management, and specialized exception handling
- •Long-run: people manage fleets of AI systems rather than do repetitive tasks
- •Maintenance labor shortage: aging workforce (many technicians 50+) makes efficiency urgent
Long-term housing future: robotics, construction costs, longevity-driven demand, and mobility
Erik asks for a 10-year view incorporating robotics and longevity. Minna connects longevity and lower cost of living to population growth and housing demand, arguing robotics/modular construction could reduce build costs and speed delivery. Tony and Minna also highlight greater mobility—shorter, more flexible leases—enabled by automated leasing and turnover.
- •Longevity + increased wealth could raise population and housing demand over time
- •Robotics/modular/manufacturing could lower construction costs and increase build speed
- •Elise is software-first today but sees hardware-enabled construction as a key missing puzzle piece
- •Mobility vision: reduced friction to move quickly and cheaply; less lock-in to 12–24 month leases
- •Automation makes frequent turnovers scalable, benefiting labor markets and quality-of-life moves
Why real estate underinvests in R&D—and why AI changes adoption dynamics
They discuss why real estate historically spent little on R&D: operations are complex, variable, and edge-case heavy, so traditional software couldn’t replace people. AI can handle larger “search spaces” and variability, making automation finally feasible and creating an incentive to digitize processes and capture institutional knowledge.
- •Real estate complexity and variability previously forced reliance on humans
- •Traditional software struggled with edge cases across different buildings and workflows
- •Overreliance on people meant critical operational data stayed in employees’ heads
- •AI can manage unstructured complexity, unlocking automation and data capture
- •Prediction: sector could jump from low R&D to high AI spend due to outsized benefits
Responding to PropTech skepticism and tackling maintenance/turn-time waste
Minna and Tony rebut the critique that PropTech only helps landlords extract more value, arguing tech typically improves consumer experience and reduces costs—especially with mass adoption and competitive pressure. They then dive into maintenance and repairs as a planning/orchestration problem where AI can reduce delays caused by missing information, bad scheduling, and poor sequencing, unlocking billions in value by shortening turns.
- •Argument: banning tech doesn’t reduce costs; tech usually increases surplus and efficiency
- •Inefficiency raises barriers to entry and can increase landlord pricing power
- •Competitive markets + mass adoption determine whether savings flow to consumers
- •Maintenance as a complex planning problem: routing, sequencing, bundling tasks, purchasing
- •Preventative maintenance and better information flow reduce delays; shaving days off turns unlocks billions
Expanding to healthcare: similar admin pain, scaling voice + scheduling, and the platform roadmap
Alex presses on the housing-to-healthcare leap; Tony explains the commonality lies in administrative workflows: repetitive inquiries, structured intake, scheduling, and phone-based unstructured interactions under heavy regulation. They argue healthcare’s rising costs are partly elastic demand and better clinical care, but admin costs have surged without a better experience—an opening for AI to improve intake, billing cycles, and post-appointment communication to boost adherence and outcomes.
- •Healthcare focus is primarily admin operations, not clinical decision-making
- •Shared patterns: repetitive inquiries, structured data capture (insurance/budget/preferences), heavy regulation
- •Voice tech and scheduling optimizations built for housing translated well to healthcare
- •Admin costs rising faster than clinical improvements; experience hasn’t improved proportionally
- •Roadmap: expand from scheduling to broader admin backend, billing cycle, and post-appointment engagement/adherence
Lessons learned and the ultimate vision: start with the underserved; cut 42% spend materially
Reflecting on hindsight, Minna says she would have started with affordable housing because it combines maximal compliance complexity with the greatest need and administrative drag. They close on the macro vision: meaningfully reduce the share of household spending consumed by housing and healthcare, making both less of a cost concern for average people.
- •Hindsight: affordable housing is most complex (compliance/paperwork) and most underserved
- •Underserved segments often have the highest administrative drag and slowest adoption
- •Strategy: solve the hardest, most constrained environment first, then expand downstream
- •Ultimate goal: drive major cost reduction so housing/healthcare aren’t primary anxieties
- •Target framing: reduce household housing+healthcare burden from ~42% to “twenty-something” percent

Publisher timestamps

Get more out of YouTube videos.

Read along on your phone, search it beside the video, or ask your favorite AI app what was said.