The Ticker May Be Too Late for AI’s Next Great Winners

Listen to the audio version of this article (generated by AI).

Back in 1957, William Shockley should have owned the future.

He had co-invented the transistor, won the Nobel Prize, and had eight of the brightest young engineers in America working under him in his Mountain View, California, laboratory.

Instead, all eight engineers quit because they found Shockley impossible to work for.

With no product and no revenue, the eight quickly realized that no institution or company would support them. Back then, the suburbs and farmland south of San Francisco and north of San Jose weren’t exactly “Silicon Valley” yet. The budding tech firms in the region weren’t quite ready to invest in unproven ideas.

So they made one phone call.

A young financier named Arthur Rock listened to their story and took a risk. 

Although Rock did not have the capital himself, he was willing to bet on people he deemed impressive.

He found a camera company willing to gamble $1.5 million on eight founders and an idea. 

Thus, Fairchild Semiconductor was born. Fairchild eventually became one of the most influential technology companies in history, spawning Intel Corp. (INTC) and dozens of other semiconductor firms worth trillions of dollars today.

And the men Shockley lost became known affectionately as the “Traitorous Eight.” They were the accidental architects of a model Silicon Valley still runs on to this day.

Source: Intel

_{The Traitorous Eight: That’s Gordon Moore – of “Moore’s law” fame – on the far left.} 

That same instinct resurfaced in 1998 when Andy Bechtolsheim  sat down with two Stanford University grad students. Right on the spot, before their company had a business model or recognizable brand, the Sun Microsystems co-founder wrote a $100,000 check to Larry Page and Sergey Brin. 

Anyone who’s ever Googled… well… anything knows how that story ended. But for the record: That $100,000 check reportedly bought roughly a 1% stake in Google, a position that eventually became worth tens of billions of dollars.

More recently, in 2023 Spark Capital invested $75 million in Anthropic while it was still an obscure AI startup with little revenue. Today, millions of people are on a first-name basis with Claude, and that stake is estimated to be worth roughly $7 billion.

Across nearly 70 years, the technologies and the players keep changing. The playbook doesn’t.

Rock backed eight unknown engineers. Bechtolsheim backed two graduate students. Spark Capital backed an AI startup few people had heard of. 

In each case, the biggest opportunity wasn’t buying a great business after everyone recognized it. It was recognizing exceptional founders and businesses before everyone else did.

I think that same playbook matters more today than it has in decades.

First, because AI has created an unprecedented race to develop new technologies. Second, because the companies leading that race increasingly have more money than time. And finally, because that combination is changing where some of the biggest fortunes in technology are being created.

Let me explain…

Why AI Giants Buy to Fill Critical Gaps

There’s a reason this playbook has endured for nearly 70 years, and it isn’t just today’s excitement over AI.

When the prize is building the next great computing platform, speed becomes everything. If a startup has already solved a problem that would take your own engineers two years to crack, buying that company is often far cheaper than losing those two years.

That’s exactly what’s happening in today’s AI race.

Alphabet Inc. (GOOG) made that decision early, back in 2014, when it acquired the British AI startup DeepMind. Rather than spending years assembling a comparable research lab from scratch, Google bought one of the world’s best AI teams outright. More than a decade later, DeepMind sits at the heart of Google’s AI strategy.

Meta Platforms Inc. (META) reached a similar conclusion last year when it invested $14.3 billion in Scale AI. The deal wasn’t just about software. Scale AI had become one of the industry’s leading providers of the high-quality training data and infrastructure needed to build advanced AI models. Instead of trying to re-create that expertise internally, Meta bought a seat at the table.

Microsoft Corp. (MSFT) made perhaps the biggest AI boom bet of all. Its $23 billion worth of investments in OpenAI, made between 2019 and 2023, gave the company immediate access to one of the world’s leading AI developers years before it could have built a comparable capability on its own.

And this isn’t unique to AI. Cisco Systems Inc. (CSCO) spent much of the 1990s building its networking empire by buying promising startups rather than reinventing technologies itself.

Long story short, this isn’t a new playbook. It’s an old one that’s becoming even more valuable.

Every one of those deals happened because the real value had already been created inside a startup, long before Wall Street ever started paying attention.

That’s why I think one of the most important shifts in investing today is this:

The buyout, not the IPO, is increasingly becoming the finish line many early investors are aiming for.