Field Notes- Faster, Better, Not Average
Over the last few weeks, I’ve had dozens of conversations about how AI is being applied inside VC firms — across stages, sectors, and geographies. It’s been a fascinating mirror of the industry itself: fast-moving, uneven, and occasionally overconfident.
Most firms jumped in head-first on the back of GPT-3. The promise of “superpowers” was too strong to ignore. Emerging managers especially were eager to move quickly in search of leverage among already skinny ranks — often deputizing a junior to “do interesting and valuable things” with AI. That speed had its benefits, but it rarely came with strategic intent. What began as exploration sometimes devolved into agents doing the thinking — and in a few cases, even distracted, inattentive behavior on calls.
More than a few people have confessed they might need a break from AI just to remember what it’s for. Not because it doesn’t work, but because we haven’t yet built the muscle of using it deliberately.
The simple framework I’ve landed on — one I’ve been using implicitly — is that AI helps us in only two ways:
- Do things faster. Automate workflows, research, and communication loops so we can spend more time on what matters — thinking deeply
- Do things better. Elevate the quality of our reasoning, pattern recognition, and creativity. I’d add it should also help us be more present with founders, our peers, and our thoughts
It’s not an excuse to outsource critical thinking. Models tend to regress to the mean — especially when context is thin and prompts are lazy. In other words: AI will meet you at the average. And that’s not how outlier investments are sourced, selected, or stewarded.
The firms that will actually win with AI won’t be the ones that automate first — they’ll be the ones that think deeper, move faster, and stay intentional.