Episode summary
Why Most Venture Capital Firms Aren't Built for AI
How I Invest With David Weisburd
· · Hosts: David Weisburd · Guests: Andre Retterath
Summary
Andre Retterath argues that venture capital can be automated in the right places, but not replaced in the parts that require judgment. Earlybird’s edge, in his view, comes from combining machine learning, proprietary data, and an increasingly systematic workflow to reduce repetitive work and improve decision-making. He says the firm built tools to surface opportunities earlier, generate better preparation, and institutionalize what individual investors learn.
A major theme is that venture is highly power-law driven, so missing the wrong deal can be far more expensive than making a mediocre one. Andre says that led Earlybird to study false negatives, build an AI-native platform, and become more flexible on issues like ownership thresholds after missing deals such as Lovable. He also argues that founder fit is highly contextual: deep tech, AI infrastructure, and application-layer startups require different founder traits, with capital attraction, talent attraction, and founder branding mattering differently depending on where a company sits in the stack.
He also says the biggest obstacle to adopting AI inside a venture firm is cultural change, not data quality. Earlybird’s internal systems, including Eagle Eye and AI-assisted memo generation, improved coverage and speed, but only after the firm tied usage to measurable outcomes like hit rate. The broader thesis is that the firms that win will be the ones that turn their own investing process into a compounding, data-driven system.
Key takeaways
- Andre Retterath says Earlybird treats AI as a way to automate venture workflows and generate alpha, not as a replacement for investor judgment.
- He says missing a breakout company can be far more expensive than backing a mediocre one because venture outcomes are so concentrated.
- Andre says Earlybird’s AI tools improved sourcing coverage to about 95% to 96% and helped the firm see 19 out of 20 relevant opportunities before close.
- He argues that founder traits are not one-size-fits-all; the best founder profile depends on whether a startup is deep tech, AI infrastructure, or application layer.
- Andre says the hardest part of adopting AI at a venture firm was cultural change, because investment professionals had to unlearn old workflows and accept new incentives.
Notable quotes
“I just googled what's the best venture capital firm in Europe, and I looked through their portfolios, I looked through the teams, and Earlybird to me was the most technical one”
“I thought I disagree, and that triggered me to do some research. So, initially I spoke to industry experts, and then later on I just started building stuff”
“about 5% of your portfolio eventually drive 95% of the outcomes. And that being said, everyone knows you cannot afford to miss these outliers”
“If you look at the underlying characteristics, it's that someone can sell you a vision”
How the episode opens
Andre, you joined Earlybird while you were still completing your PhD in machine learning. What made you join Earlybird?
To be very honest, I just googled what's the best venture capital firm in Europe, and I looked through their portfolios, I looked through the teams, and Earlybird to me was the most technical one. So, we already had back then about 80% of B2B companies in the portfolio, and being an engineer myself, I was looking for like-minded people, and the founders of Earlybird also had a PhD in aerospace engineering, industrial engineering, and so on. So, to me that felt like the right place, and I just sent one application and got in.
How did your PhD in machine learning, how does that apply to venture capital? — Everyone said you cannot automate venture capital, it's a people business, it's all qualitative data, specifically in the early-stage investing, and I thought I disagree, and that triggered me to do some research. So, initially I spoke to industry experts, and then later on I just started building stuff that helped me also throughout almost a past decade to really automate lots of the redundant, repetitive work in venture capital.
I want to get into in a bit how Earlybird's this AI-native platform, but before we get into that, I'm curious when you joined Earlybird and you joined the venture capital space, what surprised you most as an outsider? — It's a very, very close industry, and to my surprise, at at least here in Europe, most people knew each other. So, they go to the same schools, and I think it's mostly the same also in the US and other ecosystems, but I was very distant to these networks. So, I didn't know anyone when I got into the industry, and I was surprised how much and how close these networks were with the existing investors in the industry.
From the episode’s captions: roughly the first two minutes, lightly cleaned. Keep watching on YouTube →
Topics
- ai-native venture capital
- proprietary data alpha
- machine learning investment decisions
- founder branding
- venture investing judgment vs activity
Companies and tickers mentioned
- Earlybird
- Aleph Alpha
- Black Forest Labs
- DeepCode
- EthonAI
- Energy Robotics
Watch the episode
Watch “Why Most Venture Capital Firms Aren't Built for AI” on YouTube
Summary, takeaways and quote selection by Octagon. Quotes are short excerpts; the full conversation is in the episode.
More from How I Invest With David Weisburd
-
Hunter Walk: Why Venture is Getting Harder, Picking Great Founders, and Investing Against Consensus -
True Anomaly’s Even Rogers on Space Warfare, China’s Military Advantage, and the Defense Tech Boom -
Finding the Next Anthropic, Quantifying Venture Capital, and the Science of Picking Winners -
Why 95% of AI Projects Fail, Anthropic still at $2T?, and Why AI Is Still in the First Inning -
Peter Thiel's 27-Year Partner on the AI Bubble, the Venture Capital Crisis, and the PayPal Mafia -
Why Venture Is Eating Public Markets, Picking Great VCs, and the AI Boom w/Brijesh Jeevarathnam
How I Invest With David Weisburd, this episode and its artwork belong to David Weisburd. Octagon is not affiliated with the show; this page is an independent summary.
