Why AI Won’t Be the Investment Opportunity Everyone Thinks It Is — This is going to be BIG

I was watching an episode of Seinfeld the other day and it struck me how different location-aware cell phones have made our daily existence. The whole plot was based on nobody having a phone in their pocket. Seeing characters leave answering machine messages for people (and missing them), using paper maps, and getting lost on the way to a cabin made it all feel pretty dated.

Once cell phones became ubiquitous, it became much more difficult to date TV shows. Was this episode from five years ago? Ten? People had iPhones 15 years ago so you have to go back a while before a flip phone will show a show’s original air date.

When I think about true technology disruptions, the ones that enable huge outcomes for investors because they create generational behavioral change, whole new markets, and populate entire business ecosystems out of thin air, location-aware mobile devices stand out to me as a point of match the same website.

The venture asset class has already decided that artificial intelligence is the next big investment opportunity, but I’m not so sure it will disrupt business and create the global wealth that has been predicted.

In 2004, I was working for the General Motors pension fund, which had been making private equity investments in venture capital since the early 1980s. I could see all the major VCs launching their funds.

What stood out was how similar they all sounded, that is, until I got the pitch from Brad and Fred at Union Square Ventures.

It was the first time I heard someone talk about long-term cycles of disruption, not just individual technologies, and what new business models would be made possible by it.

USV introduced digitally native business models that could not exist until the Internet connected everyone. One particular example that Brad brought up was how Amazon was simply a physical store but online. Amazon’s counterpart was Sears, also a place where you could buy a lot of different things, but not anything particularly innovative.

Google, on the other hand, was a digitally native business model, with no offline analogue. It was built on cross-linked data, enabling quality search and efficient information retrieval on a massive scale.

I’ve been thinking a lot about whether AI represents the same kind of game-changing investment opportunity that comes from disruption cycles. As awesome as it is to have a conversation with a seemingly omniscient chatbot or watch it create a Wes Anderson version of Star Wars, I keep coming back to two things:

1. Most AI business models are simply “better, faster, cheaper” models – iterations of existing models currently made or that can be made by humans, but perhaps not profitably.

2. Since these models require training data, companies that have a lot of proprietary data have an obvious advantage, leaving everyone to deal with what is publicly available.

Technology has already made the world quite efficient. The idea that you’re going to get a giant Internet or mobile-sized economic opportunity in every way, better, faster, and cheaper in human-based models seems a little hard to swallow, especially when you consider the cycle of hype which is now bigger than ever. Other than early checks, no one paid a remotely reasonable risk-adjusted price for OpenAI stock unless it ends up being a FAANG-sized hit, and the same is true of many AI companies these days.

Also, the zero-sum game of the training data prerequisite appears to be a limiting factor for the asset class as a whole. While a single company may have proprietary access to a great dataset, AI innovation doesn’t seem to be a particularly level playing field. Companies like Google, Apple and Meta should have had a huge advantage in building large language models and I suspect they will catch up soon. Bloomberg announced that it had built a financial version of a language model, seemingly eliminating many early-stage investment opportunities in this space before they even started.

Also, all these better, faster, cheaper models have to be around the current ways of doing things (creating, programming, etc.), things where the stack required to complete the task is maybe 10% of AI and 90% of many things. other “table betting” parties. It’s very likely that the best place to draw with AI is still an Adobe app that uses AI plug-ins built by themselves or other companies, but those other companies won’t want to rebuild all the layers, colors, editing, and layers. other types of basic drawing tools that Adobe has perfected over the years. Again, this doesn’t seem to present a game-changing investment opportunity across the board, just as the web itself was a fairly blank business canvas when it first started.

Don’t get me wrong, there will certainly be big results for companies that integrate AI into their offerings, and we may be watching today’s sitcoms 30 years from now wondering how it turned out. going through all of this before artificial intelligence changed everything, but I’m not. sure the dollars will be there from the return side of the company.