2025 Hot Takes

What Leaders Need to Know

Earlier this week, Y Combinator’s ex partners Michael Seibel and Dalton Caldwell dropped their “2025 Hot Takes” video, a candid conversation covering everything from San Francisco’s reality vs. Twitter narratives to AI timelines, space economics, and even political messaging. For those of us in B2B tech and private markets, their discussion contained actionable signals for investors, founders, and operators.

Here’s my breakdown of what actually matters for your 2026 strategy.

1. San Francisco: The Twitter Narrative vs. The Operational Reality

What They Said:

“If you’re reading it on Twitter, there’s a greater than 50% chance it’s wrong... It’s just different. There’s no signal.”

Tech & Capital Translation: The perpetual “San Francisco is dead” versus “San Francisco is back” debate misses the point. What actually matters:

For VCs:

• Due diligence alert: Companies with “SF optionality” (distributed teams with HQ flexibility) now have structural advantages

• Portfolio strategy: Don’t over-index on geography narratives. Focus on companies that can hire/retain talent regardless of physical location

• Signal vs. noise: The SF commercial real estate collapse ≠ tech ecosystem collapse. These are different markets

For Founders:

• Recruiting leverage: The SF talent pool is still deep, but remote options widen your candidate universe

• Cost structure advantage: Companies that negotiated leases post-2023 have 30-40% rent advantages over pre-pandemic competitors

• Network effects remain: Despite Zoom, serendipitous connections still happen disproportionately in density clusters

My Take: SF isn’t “back” in the 2019 sense, it’s evolving into a high-density innovation hub without the mandatory colocation requirement. This creates optionality that smart companies are leveraging.

---

2. The Depopulation Paradox: A Looming Economic Earthquake

What They Said:

“The graphs look bad. They look really bad... I think this is one of those ideas that when they take hold, they can have a lot of power.”

Tech & Capital Translation: The Stanford Food Research Institute’s 1970s “stop having children” ideology has become today’s demographic reality. This isn’t just a societal issue—it’s an economic restructuring event.

Implications for Investors:

• Labor automation: Companies solving workforce shortages through AI/robotics become essential

• Healthcare tech: Aging populations require radically different care models (telehealth alone won’t cut it)

• Housing market transformation: Depopulation in some areas + concentration in others creates uneven real estate opportunities

• Government tech: Social security, Medicare, and pension systems need digital transformation at scale

Startup Opportunities:

1. Elder tech beyond monitoring: Actual productivity tools for aging populations

2. Immigration infrastructure: Streamlining legal immigration becomes national priority

3. Education reimagined: Lifelong learning for career transitions at scale

4. Productivity tools: Making smaller teams dramatically more effective

The Bottom Line: Depopulation is the silent restructuring force of the next decade. Companies addressing its consequences will have built-in market tailwinds.

---

3. AGI Timelines: The “Six Months Away” vs. “Generations Away” Divide

What They Said:

“I’m always skeptical of ideas that take all the oxygen out of the room... There’s so much distance between [current AI] and super intelligence.”

Tech & Capital Translation: The self-driving car analogy is perfect: In 2015, we thought ubiquitous self-driving was 2 years away. In 2025, it’s gradually happening with human oversight. The same pattern is unfolding with AGI.

Practical Implications:

For Company Building:

• Human-in-the-loop systems: Design for augmentation, not replacement, this will be the dominant model for 5-10 years

• Infrastructure over magic: The real value is in making current AI reliable, scalable, and secure

For Investment Theses:

• “AGI-dependent” bets: Companies whose business model requires AGI-level performance are dangerously speculative

• Focus on adoption gaps: Most businesses use <5% of available AI capabilities. Closing that gap is a $100B+ opportunity

• Tooling over models: While everyone watches foundation models, the tooling layer (eval, monitoring, security) is where sustainable businesses are being built

Key Insight: The companies winning in 2026 will be making today’s AI actually work for businesses.

---

4. Space Economics: The iPhone Moment for Orbit

What They Said:

“We haven’t experienced cheap launch yet... We’re going to look back at this time and be like, ‘God, that was expensive launch.’”

Tech & Capital Translation: SpaceX has done for launch costs what AWS did for compute costs. We’re at the beginning of the space-as-a-utility era.

Investment Implications:

Near-term Opportunities (1-3 years):

• Space infrastructure: Ground stations, data processing, orbital logistics

• Earth observation: Agriculture, climate monitoring, urban planning

• Satellite communications: Redundant networks for critical infrastructure

Medium-term Plays (3-7 years):

• Manufacturing in microgravity: Pharmaceuticals, materials science

• Space resource utilization: Water extraction, in-orbit manufacturing

• Space tourism infrastructure: Not just rides, but habitats and experiences

Long-term Vision (7+ years):

• Interplanetary economy: Earth-Mars supply chains

• Space-based solar power

• Asteroid mining

The SpaceX IPO Watch: When SpaceX goes public, it will create a liquid asset for space exposure, potentially driving more capital into the sector. But beware of the “SpaceX dependency risk”, companies overly reliant on one launch provider face concentration risk.

---

5. The AI Infrastructure Bubble: Debt-Fueled Capex

What They Said:

“If you look at the whole machine, there’s a lot of debt that’s being issued right now... Some of this data center stuff seems a little scary.”

Tech & Capital Translation: This is the most important insight for private market participants. We’re seeing a classic infrastructure bubble pattern:

The Pattern:

1. Hype around new technology (AI)

2. Massive capital expenditure required (GPUs, data centers)

3. Debt financing for long-duration assets

4. Overbuilding and price competition

5. Shakeout of over-leveraged players

Who Gets Hurt:

• Infrastructure builders: Companies taking debt to build data centers

• Component suppliers: Betting on sustained high demand

• Highly leveraged investors: Using debt to finance AI bets

Who Benefits:

• AI application companies: Get cheaper compute as infrastructure competes

• Customers: Better prices for AI services

• Surviving infrastructure players: Consolidate market share post-shakeout

Your Playbook:

• Avoid infrastructure debt plays: Especially those with >50% debt financing

• Focus on capital-light AI: Companies leveraging existing infrastructure

• Wait for the shakeout: Better terms will emerge when overbuilt capacity needs customers

---

6. Abundance vs. Affordability: The Political Economy of Tech

What They Said:

“Abundance is clearly the good governance... but affordability is clearly the short-term campaign message.”

Tech & Capital Translation: Tech creates abundance (cheaper goods, more services). Politics focuses on affordability (why does everything feel expensive?). This tension defines regulatory risk for tech companies.

Sector Implications:

Most Exposed:

• Housing tech: Caught between NIMBYism and affordability crises

• Healthcare tech: Price transparency vs. system complexity

• Education tech: Cost inflation meets promised democratization

Least Exposed:

• Productivity software: Clear ROI, less political scrutiny

• Infrastructure tech: Enabling abundance without direct consumer pricing

• B2B focused: Less visible, less politicized

2026 Strategy: Companies that can frame their value in affordability terms while delivering abundance will navigate regulatory environments more successfully.

---

The 2026 Action Plan

For Investors:

1. Shift from AI hype to AI adoption: The 100x returns will be in companies making AI work, not just building AI

2. Analyze debt exposure: Scrutinize portfolio companies’ infrastructure financing

3. Position for space acceleration: The SpaceX IPO will create new comps and opportunities

4. Watch demographic solutions: Depopulation isn’t reversing—back companies addressing consequences

For Founders:

1. Build capital-efficient AI: Your competitors are taking on debt—outlast them

2. Design for hybrid teams: The SF/Austin/Miami debate is over—build distributed from day one

3. Solve today’s problems: Not what AGI might solve in 5 years

4. Navigate political narratives: Frame your company as increasing affordability, not just creating abundance

For Tech Leaders:

1. Pressure test AI infrastructure plans: Is your vendor taking on too much debt?

2. Plan for talent scarcity: Automation isn’t just efficiency—it’s survival

3. Evaluate space opportunities: Not just Starlink—what data/connectivity becomes possible?

4. Monitor regulatory shifts: Affordability narratives drive policy changes

---

2025 was about the messy middle of adoption.The companies that win in 2026 will those that bridge the gap between possibility and practical implementation.