Why I'm building my next AI startup with $0 VC money
Everyone thinks you need $50M and 200 engineers to build a successful AI company.
I've raised $29M across two startups, scaled them to 23M+ users, and had Craft Ventures and Khosla Ventures write me checks.
And I'm about to prove that playbook is dead. With AI in play, the era of massive teams and huge rounds is over.
Two eras, two different lessons
Era 1: Pre-ZIRP Growth-at-All-Costs (2013)
Classic consumer startup story before anyone even knew what ZIRP was. "Growth first, revenue later" because that's what everyone preached. We hired fast, burned through millions, and nearly imploded because we couldn't figure out monetization.
But here's the plot twist—we eventually cracked profitability and turned it around. Hard-earned lesson: growth without unit economics is just expensive vanity.
Era 2: Peak ZIRP AI Boom (2020-2021)
This was completely different. We were one of the first 50 companies globally using OpenAI's APIs when most founders had never heard of GPT. High growth, high velocity, speed of execution like OpenAI themselves. We built a horizontal AI product that most companies hadn't even attempted—and scaled to 17M users with Craft Ventures and Khosla Ventures backing us.
The ZIRP context: 2021 saw $329.9 billion invested in US startups—nearly double 2020's record. Global VC activity hit $621B, more than 10x the levels from a decade earlier. Over 500 companies became unicorns in 2021 alone. Money was practically free.
We were part of that ZIRP generation. And here's the thing—not many companies survived when rates went from near-zero to 5% overnight. We're one of the few still growing.
Why AI changes everything about startup economics
The data doesn't lie: AI can boost employee productivity by 66% on average, with some roles seeing 88% gains for developers and 126% for AI-powered developers. AI adoption can drive 2.9% annual labor productivity growth—equivalent to 47 years of natural productivity gains compressed into one year.
Meanwhile, Meta nearly doubled its workforce during the pandemic (48K to 80K employees), and 476 tech companies laid off 141,467 employees in 2024 because they overhired for work AI can now do better.
You don't need 50 engineers when AI can generate code and boost developer productivity by 88%
You don't need massive data teams when AI can process 80% of unstructured data into actionable insights
You don't need huge customer success teams when AI can handle 13.8% more customer inquiries per hour
But VCs are still writing checks like it's 2021.
Why I'm going lean with Otto AI
My new company, Otto AI, is building AI back office for creative small businesses and solopreneurs. Instead of raising a Series A, I'm doing something different:
Bootstrapping first, then seedstrapping.
Not because I can't raise money—I've done it twice in the peak ZIRP era. But because I want to prove something: in the AI era, you can build massive companies with tiny teams and smart capital.
What you'll learn by following along
Every Thursday, I'll share real-time insights from building Otto AI:
Plus behind the scenes reality:
Why SMB startups that couldn't get funded for decades are suddenly raising millions (and it's not what you think)
How single developers are building companies that compete with 100-person VC-backed teams
The exact conversation I had with a top VC about why they're now funding "small" ideas they rejected in 2019
What frameworks actually work for SMB SaaS vs. the enterprise playbooks that don't scale down
Plus the raw, behind the scenes reality:
How we went from $0 to $100K ARR in 3 months without even having the actual product built yet
The teams I should have never hired vs. the teams I should have hired from day one (and why timing matters more than talent)
Why I left LA to bootstrap while traveling instead of burning cash in expensive markets
The exact playbook we're using to hit six figures before our MVP is even complete
This isn't another "follow my journey" newsletter
I'm not just documenting the ride. I'm proving a thesis:
That the smartest founders in the AI era will build lean, profitable companies while others burn through massive rounds hiring their way to bankruptcy.
The data doesn't lie: 917 tech companies laid off 144,554 workers in 2022 alone. I've been the idiot who overhired in the ZIRP era. I've been the founder who thought more money solved more problems.
Now I'm betting everything I learned the hard way plus $29M in expensive lessons from both the growth at all costs era AND the AI boom can help you skip the hiring trap.
Hit subscribe if you want the unfiltered truth about building startups in the AI era.
Building Startups by Aj Yadav publishes every Thursday. Paid subscribers get monthly deep dives with actual metrics and exclusive VC interviews.