Podcast Summary: Startup Stories – Mixergy

Episode #2281: Garry Tan – Y Combinator Startups Growing 5X Faster – Here’s What Changed

Host: Andrew Warner
Guest: Garry Tan (President and CEO, Y Combinator)
Date: October 17, 2025

Overview

In this engaging conversation, Andrew Warner interviews Garry Tan, the President and CEO of Y Combinator (YC), focusing on the explosive growth of YC startups—many now growing revenues 5x faster than ever before. The discussion explores what has changed in the startup ecosystem, the catalytic role of AI (particularly LLMs), YC’s evolving philosophy, and the new opportunities and challenges for both founders and investors in this rapidly shifting landscape.

Tone: Conversational, ambitious, and candid, with a dash of introspection and startup hustle.

Key Discussion Points & Insights

1. The Unprecedented Growth of YC Companies

• AI as a Game Changer: YC startups, even as small teams, are hitting milestones previously unimaginable in software—$10–20M ARR in 10–20 months.
  • “We’re seeing routinely YC companies with 10 or 20 people get to 10 or $20 million a year in revenue in 10 or 20 months. That’s like literally never happened before in software.” (Garry, 00:12)
• Shift in Growth Norms: Average weekly revenue growth in the YC batch has increased from 2–4% to a staggering 10–20%.
  • “It went from on average 1 to 2% to 10 to 20%.” (Garry, 12:52)
• Explosive Opportunity: Garry frames contemporary times as a “golden age” for startups—the barriers to build and scale have plummeted, and expectation levels for YC founders have soared.

2. The CaseText Story: How AI Transformed a YC Startup

• Background: CaseText aimed to modernize legal research by indexing legal documents—useful, but capped in scale.
• AI Pivot: The advent of advanced LLMs (e.g., GPT-4) transformed their offering. The critical breakthrough was learning to “chop down” legal questions into manageable, context-specific prompts, minimizing AI hallucination—a must in sensitive legal environments.
  • “If you chopped it down to a bite-sized chunk…he found that he could…have output that was usable, useful and reliable and not a hallucination, but it required you to chop that down into a particular small enough step.” (Garry, 05:41–06:41)
  • “Jake being a great designer, engineer, and lawyer…he tried really hard to make it work and it would hallucinate. And…he also was operating in an area that…could not withstand any hallucination.” (Garry, 03:10)
• Productization and Market Reaction:
  • Drastic cost savings and demonstrable product accuracy led to lawyers being “astonished,” creating a “must-buy” tool.
  • “It would cost thousands, tens of thousands of dollars [to do this work manually]…that was the feat of strength.” (Garry, 08:09)

3. Competing Against Giants & Vertical Focus

• SMB vs Big Tech: Startups can still win by deeply focusing on verticals.
  • Example: Avoca, a YC company, building tailored AI for HVAC customer support.
  • “While it’s conceivable ChatGPT or Gemini could one day do this, today it hasn’t happened yet, and there’s enormous room for vertical, purpose-built products.” (Garry, 10:00–11:55)
• The Power of Customization: Garry believes deeply customizing tools for specific industries (even “unsexy” ones) remains key, despite some predicting a future of universal AI agents.

4. Democratization of Entrepreneurship & Product Creation

• Lower Barriers, More Founders: With AI, “one engineer can do the work of 20”—unlocking software, product, and even physical goods creation for a wider range of people.
  • “Each engineer working on Claude code is doing the work of 20 people…That’s actually the good news.” (Garry, 19:06)
• Proliferation of Niche SaaS: Expect a world with “a CRM for podcasters, a CRM for HVAC, in fact, multiple of them.”
  • “The revenue of Salesforce…would then be spread across a bunch of companies, and there would be a bunch of companies whose founders are not as wealthy, but…making a strong enough living.” (Andrew, 22:12; Garry, 22:45)

5. Impact of AI on Startup Tactics & Personal Productivity

• Prompt Engineering as Craft: Both Andrew and Garry dive into the nitty-gritty of using AI for creative output. Garry shares how he uses LLMs (e.g., Gemini, ChatGPT) to rapidly iterate on high-quality video scripts for his YouTube channel by refining prompts and collaborating with the model.
  • “I could have a new 10 minute script ready…whereas it normally would take me like several hours.” (Garry, 00:27)
  • “Almost anything you rely on humans for, you could probably do better…these things are not writing for me—they’re helping me, and I’m working with it. So it’s a little bit more like a co-writer.” (Garry, 36:18)
• Willingness to Share: Garry promises to make his scriptwriting prompt public:
  • “Can we give one of these sessions to our people?” (Andrew, 35:48)
  • “Yeah, absolutely.” (Garry, 35:49)

6. How YC Has Changed Under Garry Tan

• Back to Core Values: YC has shifted from an “Alphabet” model (many arms, including a continuity fund for follow-on investments) back to a focused “Google” model, concentrating all attention on the initial batch and supporting founders for the life of the company.
  • “We explicitly decided…we’re gonna go back to being Google, and it’s easier…it’s just like, let’s do what we uniquely do the absolute best.” (Garry, 39:14)
• Long-Term Founder Support: Each founder now has a dedicated partner, forging relationships that continue for years.
  • “It’s like having your best angel investor who is there for you all the time…that’s from for many years.” (Garry, 28:07)
• Community as Superpower: The close-knit YC batch experience and alumni network continue to be a unique growth driver.

7. The Changing Demographics of Founders

• Younger, Hungrier: More YCs are now in the 18–22 age bracket (+100% YoY). Meanwhile, 25–30s are down, as those who entered during “ZIRP” hang on to tech jobs and, per Garry, are often “AI deniers.” (25:18–25:55)

8. The Role of Design and Execution in AI Startups

• Execution Still Matters: Even for products with strong AI backends (like Rosebud AI), the user experience and polish—what Andrew calls “the Garry Tan magic”—can make or break adoption.
  • “That level of detail to design needs to be applied there because the brains are good.” (Andrew, 45:15)
  • “That could be fixed tomorrow.” (Garry, 45:23)

9. Advice & Reality Check on Building on Top of AI Platforms

• MVPs vs. Real Businesses: Simple “wrappers” on consumer AI APIs may demonstrate opportunity, but true defensibility and growth still require strong engineering and iterative product refinement.
  • “Ideally the founders themselves are actually really cracked engineers and then you basically 20x yourself…If you actually learn how to prompt…you’d be a 200x engineer.” (Garry, 47:14)

10. Vision for the Future

• Consumer AI Isn't Done Yet: Garry sees room for consumer-facing innovations, especially those replicating high-value human services (therapy, coaching) at scale and accessibility.
• Abundance & Societal Change: If AI achieves superintelligence, “all bets are off”—but in the meantime, abundance for founders and customers is within reach through entrepreneurship in every niche.

Notable Quotes & Memorable Moments

• On why now is a great time for startups:
  • “This is by far the best time in the history of startups to be starting one.” (Garry, 47:50)
• On prompt engineering:
  • “The cool thing about prompts is that it’s actually an intelligible version of fine tuning the model.” (Garry, 30:56)
• On founder mindset:
  • “The most important thing actually is that you don’t really know who to trust. So if you have at least one person at YC…that’s actually better than I think 90% of people who start startups, period.” (Garry, 41:24)
• On market opportunities:
  • “Infinity sort of off the beaten path, like underserved verticals. Like HVAC is just one of like…there’s beer, there’s gold.” (Garry, 42:28)
• On the shift within YC itself:
  • “It’s a thousand times easier to be Google than to make Alphabet work.” (Garry, 39:14)

Timestamps for Key Segments

| Theme/Topic | Timestamp | |---------------------------------------------|-------------| | AI & YC company growth statistics | 00:12–02:00 | | CaseText’s AI transformation | 03:10–08:37 | | Custom vertical SaaS vs Big Tech | 10:00–13:45 | | The rise of niche products | 19:06–22:45 | | YC’s new/old focus under Garry’s leadership | 37:38–41:02 | | Founder support & alumni network | 27:12–30:32 | | Prompt engineering for content creation | 30:56–35:48 | | The next wave: underserved verticals, consumer AI | 42:28–45:15 | | Building on AI platforms, MVP vs defensibility | 46:44–47:50 |

Conclusion

This episode offers an insightful look at how Y Combinator and its startups are riding a “once-in-a-generation” wave of AI opportunity. Garry Tan’s perspective highlights both the practicalities—customization, speed, disciplined execution—and the inspiring possibilities for a new cohort of young, ambitious builders. Above all, the episode is a testament to how startup success still requires focus, iteration, and a deep understanding of real customer needs—even as the tools change at warp speed.