FOMO and AI
I first heard the word "FOMO” while walking the halls of Stanford Business School about 15 years ago. It was used to explain behavior like signing up for too high a course load, casually “surfing” multiple campus activities at the same time, and even the selection of roommates for a particular study trip.
By now, you probably also heard of the term; FOMO stands for Fear Of Missing Out. It is that feeling that somewhere else something better is happening and you’re out of the loop.
Over the past decade, FOMO has been exported to all corners of the world, courtesy of social media. It is probably no coincidence that FOMO and Facebook were born around the same time (2004) and in the same location (Harvard), but it was Silicon Valley that ran with both – and never looked back.
FOMO can explain many behaviors in Silicon Valley. For example, it is rampant in angel investing. We’re always looking to discover the next Google - and get in on the ground floor with valuations we can afford. FOMO also drives the coffee meeting concept. Who knows if that humble grad student that wants to meet me over coffee may become the next Sergey Brin? The VC community is obsessed with FOMO for a slightly different reason – they need to impress their limited partners. It is easy to forget – VCs don’t invest their own money; they get their funding from limited partners. So, whenever there is a new technology trend, the VCs will rush in; nobody wants to have to tell their bosses that they failed to spot and invest in the hottest new trend. Ironically this first VC rush will cause even more FOMO among those that haven’t committed yet, and the inevitable lemming effect takes hold…
FOMO explains the bulk of current investment dollars going into the AI sector. Forget about building MVPs (Minimum Viable Products) on a shoestring budget, the challenge now is scaling up faster than your competitor. That’s a bit of a problem if that competitor is Google, which already has a $140B data center infrastructure in place. So, if you’re OpenAI that means you need to announce a $500B plan to quickly build out your own infrastructure. But Google does not want to become the next Yahoo, so it responds with its own $100B a year plan to rapidly expand its own infrastructure and soon, we will be talking in trillions of dollars…
The current approach to reaching the coveted AGI (artificial general intelligence) milestone seems too brute force to me (and, perhaps more importantly, to others as well). Just like LLMs (Large Language Models) came out of nowhere to bring us intelligent chatbots, I have a feeling that some other new concept will need to emerge to realize AGI. And yes, there is a host of “long-tail” startups right now building on top of LLM technology to reinvent business processes, heralding a new age of efficiency. But to improve their margins, those startups will eventually build on top of free open-source LLMs which are nearly as good as the latest models from OpenAI, destroying a key part of the AI behemoth’s business model. You can grow revenue only so far on $20/month consumer subscriptions, especially if the “free” tier is almost as good…
While FOMO was officially coined in 2004, the behavior was of course also prevalent in prior boom/bust cycles, such as the dotcom bubble in 1999. Should we fear the “inevitable” AI bust? If the dotcom crash is any guide, we might see some good emerge after the initial gnashing of teeth. As the huge dotcom companies started going bankrupt starting in 2001, server and telecom infrastructure became available at fire sale prices. This heralded the age of free long-distance phone calls, free music streaming (and later videos), and the rise of SaaS (Software as a Service) startups. What can we look forward to, by say 2030? Maybe, instead of free phone calls, we will get free electricity to drive our EVs and run our heat pumps – or at least it will be included in our Xfinity subscription… Any better predictions? I am all ears!
As you can surmise by now, I try not to let FOMO drive my decision making. As a wise AI (the free tier of ChatGPT) once said, “In tech, fear of missing out creates noise; conviction creates companies.”