Notice: Your Investment Committee Has Been Replaced by AI

You know how, when companies send out press releases explaining why key executives are no longer in their jobs, the press releases always say the person “left to pursue other interests” instead of saying they were fired or laid off?  Well, given what’s going on in the world of private equity/venture capital and AI, I think that standard press release quote may soon change to “replaced by AI, hope they have other interests.”

“A” and “I” were always popular among vowel enthusiasts, but now, when used together, they are rapidly combining to become the most commonly-used word on the planet, recently replacing “the,” “and,” and “cocktail” as the most used words in any after work setting.  And during work settings.  And in pre-work settings.  Sorry about my cute and anachronistic overuse of the word “and.”  I will start replacing that with AI shortly.

This week I heard about and even found actual articles (through my Google AI agent) about how a few private equity firms are “putting AI on their investment committees” as voting members.  In other words, they are not using AI for analysis alone but giving AI status as a voting person alongside the usual suspects who previously voted on such matters.  See these two stats found here and here, for instance.

The survey done at the British Venture Capital Association event found that “48% of the audience said that they expect AI to be a voting member of their investment committees within the next five years.”  I wonder if that 48% is aspirational or a cry for help?  Is it just a British thing or will we in America seek to win the AI instead of investor humans arms race?  At least in Britain you get a nice cuppa tea on your way out of the committee room.

You may say to yourself something like this: PE/VC executives and AI have similar personalities in that they act like or at least believe they know everything, so what’s the difference between me, my colleagues and AI?  The difference is that AI doesn’t need to get any carried interest.  So that’s a win for the remaining humans who share the carry unless the AI figures out that it’s not making enough cash in exchange for its contributions to returns.  That could immediately start a  Hal-like messaging cascade about the AI’s plan to withdraw from the partnership unless it gets compensated correctly.  This ought to be interesting.

I can’t help but wonder why using AI for an investor vote is a good idea, at least in the current world.  First of all, fund partners are already known to draw conclusions in the same manner that AI hallucinates, so what is the difference in expected outcome exactly?

Also, AI is fed by data from the past.  While it can draw conclusions from what it knows, could it see actual transformative innovations coming?  Would it have predicted the existence of the business models underlying Uber? Airbnb? Waymo? Starbucks in its day?  Would it have predicted the long-term impact of COVID on all businesses before they happened?  It’s hard to imagine how AI could dream up new business models that have never been tried and evaluate business plans that seem crazy in the moment but are actually “yep, that’s the world now” 10 years later.

I can imagine a world where AI could apply known business models to industries that have yet to be impacted by them, so maybe that’s the reason it could theoretically work in PE but less so in venture capital funds that prefer swing for the fences ideas to incremental innovation.

I have a hard time seeing an AI bot standing on stage at any number of venture/PE conferences as it is very hard to fit a fleece vest on a piece of software, though maybe that’s the business idea of the century!  Off to start my new company, which creates bespoke virtual fleece vests for software in the cloud; I have dubbed my newco  “Fleeced.”^TM

One thing that allegedly is pleasing to the investors who provide the money to support venture and private equity funds is that the so-called human partners making investment decisions have lived experience, pattern recognition and the creativity to serve as quasi-futurists – all of that has historically mattered to most investors, as has investment track record.  Is that all replaceable with AI?  Do those people who raise the funds end up serving the AI in the end instead of vice versa?  Is our new robot overlord a distributed computing creature and we, the fund employees/partners, merely the order-takers of the future, collecting capital for our Senior Managing Partner, Dr. A.I. Cloud, to invest on behalf of its financial supporters

The funny thing about applying data to abstract circumstances and prediction is that sometimes it works and sometimes it doesn’t.  If, in the end, the companies themselves are still run by humans who do things that aren’t entirely predictable (both good and bad), how can AI bots see that in their models?  It’s certainly true that only one half of venture/PE investors have results better than average, so maybe that’s good enough, but the prospects for a long-term collective vacation for me and all of my investor friends sounds both terrifying and awesome at the same time.

If you’re a fan of the movie Space Odyssey 2021, there are so many quotable quotes for this moment.  If you haven’t seen the movie, the special effects now look lame, but the story line is right out of tomorrow’s news.  Check it out.  Note that HAL is the AI on the spacecraft, as your first clue. HAL stands for “Heuristically-programmed ALgorithmic Computer.” If you think AI is a new idea, think again.  This movie takes place more than 50 years after the original concept of AI came forward from Alan Turing and others.  So hey, maybe we aren’t moving fast to anywhere.  We are just trying to break things.

A few examples from Space Odyssey:

Imagine for a moment that Dave, HAL’s costar in the movie, is the existing human on the investment committee…I like this example because the odds of the lead investment committee member being named Dave in the average meeting are pretty high.

How it sounded in the movie:

• Dave Bowman: Open the pod bay doors, HAL.
• HAL: I’m sorry, Dave. I’m afraid I can’t do that.

How it sounds in the investment committee discussion

• Dave Bowman: I want you to vote for my deal, HAL.
• HAL: I’m sorry, Dave. I’m afraid I can’t do that.

And this below will be perfect for deals that fail because there are still some humans on the investment committee, not to mention running the company:

• HAL: [Regarding the supposed failure of the parabolic antenna on the ship, which HAL himself falsified] It can only be attributable to human error.

So, hey investor friends, I get it.  The AI pull is so strong and perhaps it is impossible to beat it so we must join it.  But just remember.  The data driving this train came from…us.  Ceding power to the AI machine may just make it harder to keep yourselves in fleece vests and may not result in better fund returns.  Let’s step cautiously into this one.  AI can be a great tool, but should it be our pink slip in a field that has been touted for individual vision?  Then again, work is hard and there’s nothing wrong with a cocktail.

I regret to inform me that I just realized that you can’t spell cocktail without “ai.”  There is no escape.