The Hype Train Has Left the Station

It’s quite tempting to be deceived by a name, and the business world has been doing it for years. Back at the turn of the century everything with .com was instantly more valuable. In the 1960s, consumer products lent into the space race. Now, every up and coming start up has AI in its name. I’d like to remind all the branding and marketing folk that just because Torquay is on the so-called English Riviera, it doesn’t make it as splendid as Portofino. It’s still Torquay. 

None of this is to say that the AI era will not be anything other than huge technological and market change, ushering in huge companies, changing the world, adding to GDP and giving customers better outcomes. We know that was the case in the dotcom boom, where the bust resulted from too much hype too soon, but not misplaced hype. Most of what was predicted during the dotcom era, such as social media, seamless online payments and financial services, delivery services, booking taxis over the internet and quickly delivered groceries, happened. It just took a bit longer than envisioned because it needed the development of the smartphone and the app stores. As Fred Wilson says, “Nothing important has ever been built without irrational exuberance.” He might very well be right: tulip mania, the goldrush, the steam engine and the proliferation of UK railways, the dotcom boom (and bust), and now, AI hype.

We don’t yet know the outcome of the AI era, but we see the direction: vastly more consumer applications from AI; and enterprise processes will be augmented and automated completely by artificial intelligence software. 

It’s easy to be taken in by the spin. How do you know which startups will be the Big Tech of the AI era? Which will be The Big AI. Is it possible to predict? I know of more than one company claiming to use AI when it’s actually manual, human work. To avoid being sued, I ain’t going to name those companies.

Everyone promises to change the world before they’ve shipped a product. Swooning VCs go big on delirious ambition without looking at the operational details. And yes, I’m looking at you, Sequoia Capital and SBF (the internet never forgets). A vision of an app that can do everything, even “buy a banana”, that’s the ticket! Let’s write a cheque! 

This brings me to my second point. I’ve long believed that strategy consultants and departments are overrated and bordering on pointless. It does not take a thirty two page PowerPoint presentation to describe to me, or for that matter, anyone else, that AI will change things. I do not need a multimillion-pound consultant to tell me, or anyone else, to cut costs and grow revenue.

Every company that ever failed had a strategy. Yes, a bad strategy is sufficient to fail. However, a good strategy is not sufficient to succeed. It’s necessary, but not enough. Because what sets the winners apart from the losers is the detail, the tactics and the operational effectiveness. It’s the execution. It’s why Amazon is Amazon, and Pets.com is a relic from the dotcom bust. Bezos had a humongous ambition to dominate markets, but he knew that you need to be at the operational coalface to make it work. The detail matters. The customers matter. 

Ambition is a pre-requisite, but ambition without execution is nothing. Otherwise, we’d all be as rich as Warren Buffett.

So, how do you survive the “AI startup change the world strategy” hype cycle? The answer is to have the inquisitiveness of a child. You know the type, the ones that always ask why. And then ask why again. And again. And again. Annoying but frighteningly effective. The simple questions always are. It’s how the massive floor in Bitcoin could be exposed: is it cash equivalent, a stable store of value, or an appreciating asset? The best question in business is, “how do you make money”? It’s what Bethany McLean asked about Enron and got wind of the fraud (“How exactly does Enron make its money? Details are hard to come by…”)

Asking the simple question is not completely foolproof, but it’s pretty close. Some AI startups will probably change the world, and you need to be prepared to figure out which.

They used to say that José Mourinho, in his pomp, could describe how a match’s tactics would evolve before kickoff. There’s a reason successful football managers are on the touchline during a game, not in a corporate HQ reading a report from Bain. 

There’ll always be surprises, but the same should be true in business. AI is one of those surprises; no one expected that “creative” work like writing, drawing or movie creation would be one of the first industries to be truly disrupted by AI. Similarly, in the year 2000, no one knew that AWS would power Amazon’s monumental growth. But even then, in late 1999, Marc Andreessen and Ben Horowitz founded Loudcloud (later renamed to Opsware), which was one of the first companies to think about cloud computing and software-as-a-service. Cloud computing was coming. It just wasn’t clear who was bringing it.

Interestingly, it’s often not those developing the technology who are best placed to answer this. Steve Jobs and Jack Ma, two of the greatest technologists of all time, didn’t code a single line between them. But they both knew how to change the world through technology. The great John von Neumann, the greatest scientific mind of all time, supported a preventive nuclear strike on the USSR. Another incredible scientist, J. Robert Oppenheimer, was supportive of international control of nuclear weapons, which would have meant sharing nuclear knowledge with the USSR. That’s not to disparage either of these two titans of science and history but is instead to show that those at the technological coalface are not the ones at the economics coalface. 

Can the person claiming to be “changing the world through AI” accurately and probabilistically describe the market dynamics and economic forces that will lead to AI changing the world? Can the person explain the market levers and consumer behaviour leading to changed markets? This is not to be confused with a cost-benefit analysis of highly uncertain innovations. No one could have accurately calculated the full benefits of technologies like the Shinkansen, the internet, the modern production line, cloud computing, driverless cars, or the smartphone. Absolutely no one. Cost-benefit analysis is a mechanism for preventing investment and should be avoided at all costs.

But back to my question: Can those claiming to be “changing the world through AI” accurately and probabilistically describe the market dynamics and economic forces? 

If yes, you might be onto a winner; if not, then run away.