This feels a bit business-school-y to say ... but broadly speaking, new technology can be used in one of a few ways:
1. Strategy enhancement -- the technology doubles-down on existing moats and rights to win
2. Create new strengths -- expand the business in new directions
3. Operational efficiency -- make admin tasks quicker
4. Required technology adoption -- must-have technology just to keep up the times
I've been spending more and more time with different ideas and strategies on how to adopt AI well at our investment shop. We've had the chance to talk with a few other investment firms (as well as dive into other folks' LinkedIn posts, hear start-up pitches, hear vendor pitches, etc.)
From what I've seen, I believe that the most successful AI adopters will fill up bucket #1 -- AI will extend an existing (winning) strategy even further. This would be a useless post without a few examples:
1. One AI-forward venture shop's core strategy is early-early-stage AI start-ups, targeting companies before they're necessarily even formed. They built tools (with AI) that allows them to (a) read, digest, and map the latest computer science papers and (b) track and rank people (e.g. X person left Y company, another person's stock is likely fully vested on Z months after being acquired, etc.). It feels like they are a step ahead of the masses (and a small step ahead of tools like Harmonic), allowing them to get a narrow lead in sourcing.
2. I've also built a few tools (using AI) for myself that doubles down on our core strategy at Connecticut Innovations. Our biggest constraint is that companies have a presence in the state, so I found a source (CT Business Registry) that gives us a good jumping off point. Layer in a few other AI assistance, and this has become a renewed source of sourcing alpha.
3. One much more contentious area is investment memo writing (of which I've written about before). I've seen venture shops with great exited companies have lengthy, almost unreadably-long memos generated with a heavy hand of AI. I'm sure there are also great, principled memos written about losing start-ups.
Through the above lens, though, the question changes from "should we use AI to write memos" to "is writing an investment memo a core part of our investment strategy." I think for some places (like the Yale Endowment), to write is to think (a sentiment that I espouse). But, perhaps, the difference is one of strategy, not capability -- they intend to be some of the most patient capital, really hoping to partner over the next 20+ years. If it takes an extra 6-12 months to do diligence (and to write a memo and grapple with minutiae and diction and brevity), then so be it.
Venture, especially now, may be different. In some AI deals, there is no time to think or do deep deep diligence; you must see and believe the uber-high-growth story. Perhaps here AI memos are good. Or perhaps for a single GP whose core strength is their network and business building, memos are a necessary evil, an artifact produced out of necessity (but not part of the alpha). Through this lens, memos are operational, not a strategy.
I, of course, disagree, especially where I am today. I'd like to be a good investor (not just a good venture investor), and so the painstaking writing is worth it.
No comments:
Post a Comment