Startup Stories – Mixergy

Episode #2297: Elad Gil backed 40 unicorns. This is next.

Host: Andrew Warner
Guest: Elad Gil
Date: February 24, 2026

Overview

This episode dives deep into the changing landscape of startups, AI-driven business opportunities, and the future of software creation, all through the lens of legendary investor and entrepreneur Elad Gil. With a track record that includes backing 40 unicorns (Airbnb, Coinbase, Stripe, OpenAI, etc.), Elad’s perspective is both practical and visionary. The discussion explores the evolution of what qualifies as an impactful startup, the debate over “toy” projects, and where the biggest opportunities are being built today.

Key Discussion Points & Insights

1. Elad’s Track Record, Self-Assessment, and Perspective

Timestamps: 00:00–03:44

• Notable companies backed: Anduril, Airbnb, Coinbase, Instacart, Stripe, Square, SpaceX, OpenAI, Perplexity, Cognition, and others.
• On ‘making it’:
  • Elad reveals he still doesn’t feel he’s “made it.” It’s always relative, especially when surrounded by overachievers.
  • “I've never felt like I've done enough... I was like 16 and I was having an existential crisis of, like, 'I'm 16, I haven't done anything yet.'” – Elad Gil [02:09]
• Inspiration from peers: He cites Patrick Collison (Stripe) and Brian Armstrong (Coinbase) for venturing into ambitious, world-impacting projects like biology institutes and anti-aging.

2. Announcements, Art, and Societal Projects

Timestamps: 03:03–03:44

• Elad hints at projects focused on large-scale societal beauty and art (“statue or monument”), though not ready to announce specifics.
  • “Our hope is to announce our first sort of statue or monument soon...for societal beauty, like art and society at large scale.” – Elad Gil [03:13]

3. “Toys” and the Path from Side Project to Unicorn

Timestamps: 03:44–09:41

• Paul Graham’s “toys” philosophy:
  • Elad expands on how many successful products started as simple “toys” or hobbyist projects (Twitter, Airbnb, Minecraft).
  • Consumer internet “toy” era has faded; today, SaaS buyers expect fully baked solutions.
• Redefining “toy” projects:
  • He distinguishes between “toy-like” (dismissed as silly) and genuine experiments or MVPs built to solve real problems.
  • “Almost any really early model before you really scale, it feels like a toy... But it'll give you a hint or glimpse of the future.” – Elad Gil [04:18]
• Example:
  • Real estate broker builds a call-automation tool for his own pain point—Elad classifies this as “building for yourself,” not a toy.
• AI’s democratizing force:
  • New AI-enabled tools empower millions to build things, expanding who can create the “next Minecraft.”
  • “It's going to put the power of building things in the hands of millions of people who couldn't do it before. It's already doing that.” – Elad Gil [08:30]

4. Scale, Market Size, and Startup Outcomes

Timestamps: 09:41–12:55

• Big companies get bigger:
  • The largest companies (Microsoft, Google, Nvidia, Meta) are now much larger than ever. “Aggregation” at the top continues.
  • Still, smaller projects can now reach $500M+ instead of being capped at $100M.
  • “The biggest things in the world are only getting bigger. ...But it also means that a thousand flowers can bloom and instead of being $100M companies, they could be $500M companies.” – Elad Gil [09:56]
• Venture environments:
  • Silicon Valley pushes ambitious capital-fueled growth, while elsewhere there's pressure to get profitable smaller, often reducing ultimate impact.

5. Custom Software, Agency, and the Business of AI-Enabled Dev Shops

Timestamps: 12:55–16:32

• AI dev shops:
  • Companies like Brinko (incubated by Elad’s group) are building AI-powered solutions for large enterprises, accelerating digital transformation in government and big business.
• Example:
  • Automated government permit approval: Months of manual work reduced to hours using AI, then generalized to other permitting workflows.
  • “That'd be an example… we built a permitting approval flow for a government...a few months of manual review...condensed into like an hour.” – Elad Gil [14:52]
• Future of software:
  • More rapid customization atop existing data platforms (like CRMs and ERPs).
  • Questions remain whether legacy SaaS platforms will survive or be replaced by custom AI solutions.

6. The Evolution of Off-the-Shelf Software

Timestamps: 16:32–21:01

• Off-the-shelf vs. bespoke:
  • Legacy solutions persist if cheap, secure, and well-integrated; otherwise, AI coding agents may drive more customization.
  • Most companies deploy their best talent on product, not internal tool rebuilds—future “coding agents” will force prioritization of what gets replaced.

7. The SaaS Business Model – From Per-Seat to Per-Labor-Unit

Timestamps: 21:01–24:24

• Unit economics shift:
  • Move from “per-seat” SaaS to “per-unit of labor” (e.g. metered usage for agentic software).
  • Example: Decagon (AI-powered customer support) replaces recurring software subscriptions with labor cost equivalence.
  • “You're shifting labor sources in some sense over to more and more AI...that to me is a really fundamental shift.” – Elad Gil [24:24]

8. The Dual Future: Custom Software vs. Repeatable SaaS

Timestamps: 24:24–27:14

• Custom/bespoke AI tools will continue, alongside scalable SaaS.
• Large clients (governments, enterprise) have long commissioned custom builds; with AI, this accelerates.
• Vibrant opportunity space remains for one-person startups and teams building both highly specific and broadly reusable solutions.
  • “There's going to be way more custom stuff, way more creators, way more people building stuff. I think that's great, that's good for the world.” – Elad Gil [27:14]

9. Elad’s Current Projects & Behind-the-Scenes Work

Timestamps: 27:14–31:59

• Alexandria Project:
  • Working with OpenAI, Anthropic, and others to translate and audio-encode the world’s most important books into all major languages.
  • “Creating audio versions of [books] in every language…and...providing it as a free resource on the Internet.” – Elad Gil [27:35]
• Personal workflow:
  • Heavy use of AI tools (Claude, OpenAI, DeepResearch) for scraping, analyzing, and normalizing large datasets (e.g., ADHD diagnosis trends).
  • “I've been doing a lot of go scrape this information and let's start interrogating it.” – Elad Gil [28:44]

10. Philanthropy, Art, and Tech’s Societal Impact

Timestamps: 31:59–33:29

• Previous generations left physical monuments and art; tech billionaires have focused on foundations and social change with mixed results.
• Elad believes attention is shifting (or should shift) toward supporting public beauty and art alongside, or instead of, (sometimes misguided) social engineering.

11. If Elad Were Starting Out: Where He’d Build Now

Timestamps: 33:29–36:48

• Several promising areas:
  • Foundation models for materials, physics, bio, etc. (“reinterpret physics through the lens of a model”)
  • Application-layer AI—especially in consumer, where there’s still a “dearth” of new products
  • Running incubator programs to spark novel uses for new models
  • “There's so much stuff to do. It's such an exciting era.” – Elad Gil [35:09]

12. Incubation and Company Transformation with AI

Timestamps: 35:45–37:41

• Sometimes his group considers buying existing companies and transforming them with AI—especially where there’s a customer base, but legacy tech.
• Example: Revamping travel agencies with AI-driven products (Navon is cited as a company doing some of this).

13. Waves and Cycles in Tech Investing

Timestamps: 37:41–40:19

• “The obvious thing is what you should do,” says Elad—AI is still at the beginning of its major growth phase, just as social, SaaS, and fintech each took a decade to mature.
• Investors/founders shouldn’t rush to look for “what’s next” before the current wave plays out.
  • “For the next decade plus, there's going to be really, really interesting things happening. ...Basically everything, software is AI.” – Elad Gil [39:38]

14. Consumer Tech Today – Prospects and Wedges

Timestamps: 41:12–43:27

• Elad’s consumer tech investing history includes Airbnb, Instacart, and more.
• The line between consumer and “pro-sumer” is blurring (e.g., Midjourney, Perplexity, OpenAI).
• Massive incumbents now dominate distribution—tougher for startups, but opportunities exist for those who can wedge in on “ignored” problems.
• Vinod Khosla cited: “Your market entry strategy is different from your market disruption strategy.”

15. Personal Side & Closing

Timestamps: 44:27–45:10

• Outside of work: hiking, working out, yoga, travel, and snowboarding.
• Playful banter between Andrew and Elad about leisure activities.

Notable Quotes

• “People always say, oh, when's the first billion dollar single person company? That was Minecraft. That was over a decade ago.” – Elad Gil [00:00/08:30]
• “I've never felt like I've done enough. ...You see what people are doing and you're like, I want to do more.” – Elad Gil [02:09]
• “If you look at the consumer internet wave, those things truly felt like toys... and then they ended up becoming big companies.” – Elad Gil [04:18]
• “It's going to put the power of building things in the hands of millions of people who couldn't do it before. It's already doing that.” – Elad Gil [08:30]
• “The biggest things in the world are only getting bigger. ...And the question is, does that continue or not? And I don't see why it doesn't.” – Elad Gil [09:56]
• “You're shifting labor sources in some sense over to more and more AI and that's going to happen throughout the whole services world and services economy.” – Elad Gil [24:27]
• “There's going to be way more custom stuff, way more creators, way more people building stuff. I think that's great, that's good for the world.” – Elad Gil [27:14]
• “Creating audio versions of [books] in every language…and...providing it as a free resource on the Internet.” – Elad Gil [27:35]
• “For the next decade plus, there's going to be really, really interesting things happening. ...Basically everything, software is AI.” – Elad Gil [39:38]

Summary Table: Important Segments

| Timestamp | Topic | |------------|-------------------------------------------------------| | 00:00-03:44| Elad’s track record, self-perception | | 03:44-09:41| The “toy” phase in building companies | | 09:41-12:55| Market size evolution, big getting bigger | | 12:55-16:32| The business of custom AI development (Brinko, etc.) | | 16:32-21:01| Off-the-shelf vs. custom/AI-generated software | | 21:01-24:24| SaaS business model changes (per-labor-unit) | | 27:14-31:59| Elad’s current projects (Alexandria, data analysis) | | 37:41-40:19| AI as “the” wave; don’t look beyond too quickly | | 41:12-43:27| Consumer tech, company “wedge” strategies |

Episode Tone

The episode is candid, insightful, and forward-looking. Elad Gil maintains a humble yet analytical demeanor, emphasizing both grand ambition and optimism for small-scale innovation. Andrew Warner elicits practical examples and pushes Elad to clarify how one-person and niche projects can still win big in the AI-powered future.

For First-Time Listeners

Whether you’re a founder, investor, or just startup-curious, this episode is packed with accessible wisdom about building the next big thing, the changing definition of “startup,” and where the frontier is for creators—no matter your starting point.