Anthropic's 1st Round: We Got 21 No's

By 20VC

Categories: VC, Startup

Summary

Anthropic faced 21 rejections in their first funding round because VCs couldn't grasp the breakthrough technology—even those who created GPT-3. The founders repositioned from a $500M ask to a $100M seed round, eventually securing a $4B partnership with Amazon by making their value proposition legible to strategic partners who understood the AI landscape.

Key Takeaways

  1. VCs rejected Anthropic despite founders' pedigree because they couldn't evaluate the technology itself. This reveals a critical gap: when pitching deep tech, you must educate investors on the fundamental breakthrough, not just the business model.
  2. Anthropic reanchored from a $500M seed round to $100M—a massive repositioning. This tactical shift shows founders sometimes need to lower asks to match investor appetite, even when building something transformative.
  3. Amazon became the strategic buyer where VCs couldn't see the value. They understood Anthropic's positioning as delivering 'a unit of intelligence for six times less' because they were tracking Azure and OpenAI. Target informed strategic partners, not just generalist VCs.
  4. Founder context matters: having ML grad school training let the founder navigate educating investors on cutting-edge breakthroughs. Deep technical credibility is a founder asset when selling moonshot ideas VCs struggle to comprehend.
  5. The 21 rejections underscore that deep tech fundraising is fundamentally different—you're not just selling a business, you're establishing proof of concept and educating a market that may not yet understand the innovation.

Topics

Transcript Excerpt

We got 21 nos, right? And I was like, what are you guys thinking? They said, well, an this recipe sounds good in theory, but like where's the proof? These are the guys who invented GPT3. How much more proof do you want? And they said, what's GPT3? Oh my god. How do you go about educating somebody who doesn't even understand the technology and the breakthroughs that are happening in the machine learning community? Now, I was lucky cuz I had that training from grad school. I'd start a computer vis...