If Artificial Intelligence is set to wipe out journalists’ livelihoods, then the hacks are making hay while the sun shines. Because I must have read a hundred articles predicting journalism’s demise since OpenAI released its AI chatbot, ChatGPT.
And while it’s seemingly still obligatory to have the robot wordsmith knock out a few paragraphs of any article about AI as a wheeze, for the most part it continues to be journalists writing these pieces – and so far predicting the demise of their gainful employment rather more than they’re actually losing their jobs.
Yet I do understand the fear-mongering. ChatGPT has also put the willies up me, as somebody who mostly makes a living by stringing sentences together.
My designer friends were similarly freaked out by the recent Adobe Photoshop demo. This simultaneously wowed them with the power of its generative AI tools – while implicitly showing a bright six-year old could now do most of their day job.
From accountants to lawyers, actuaries to quants, white-collar workers of the world are united in fretting that AI is about to do for brain-work what word processors did for the typing pool.
Buy buy AI
To be honest, predicting how or even whether AI will press the reset on human civilisation is a bit above my pay grade.
But I do know that if everything is going to be turned upside down by brains-on-demand, then as an investor I need to own a piece of it.
That’s easier said than done, if you mean to beat the market too.
As always there will losers as well as winners from any upending of the world order. Picking between them – without overpaying for the no-brainers – will be the hard part.
Also, let’s face it “If you can’t beat ‘em, buy ‘em” is not a conclusion that demands an AI super-intelligence to reach.
That’s why there’s already an AI investing frenzy underway. A mania that has arguably inflated the most accessible plays on AI to bubbly levels in short order.
For instance, US fab-less chip designer Nvidia became the fifth most highly valued company in the world last week after it forecast barnstorming sales pinned entirely on AI-related demand for its processors, as tech giants scramble to expand the data centres required to ‘train’ ChatGPT-style AI language models.
Management lifted expectations for next quarter’s revenues by 52% to $11 billion. That’s the kind of jump you expect to see with a small cap finding its feet. Not a 30-year old veteran of a half-dozen previous tech cycles like Nvidia.
Giddy traders sent Nvidia’s shares up 24% in a day – enough to lift its market cap by more than $180bn to over $960bn.
For context, Nvidia’s erstwhile rival and former chip titan Intel’s market cap is a mere $120bn.
You could say nVidia’s shares by rose one-and-a-half Intels…
A high price for hope
Does that kind of rise – and a near $1-trillion market cap – seem right to you?
On the one hand, we shouldn’t anchor to Intel’s erstwhile position in the tech ecosystem. The sector abounds with the almost-forgotten giants of yesteryear.
Yet Intel’s trailing 12-month sales are still more than double that of Nvidia’s.
Absolutely, Intel is not growing at more than 50% quarter-over-quarter. But how sure are you that will continue for long at Nvidia?
You’d better be very sure if you own the stock, given Nvidia is trading at a nearly 40-times price-to-sales ratio.
Sky-high ratings like that were last seen sported by the most fashionable cloud tech players in 2021 – shortly before they crashed to earth in 2022. And I’d respectfully suggest Nvidia’s business is fundamentally less attractive, more cyclical, and less scalable than software-as-a-service.
Don’t get me wrong, it’s hard to think of a more impressive company right now than Nvidia. Nor one more immediately exposed to the growth of AI.
But I’m sure everyone thinks the same thing. Hence it’s pretty likely that most of its future growth is already in the price.
Tech bubble: 3.0
There’s more to this hype cycle than Nvidia. Step back a bit, and you can make the case AI mania is the chief driver of the advance of the US market in 2023.
The S&P 500’s rise year-to-date is largely due to gains from Nvidia and the other tech behemoths: Apple, Microsoft, Alphabet, Meta, and Amazon. (Amazon is officially a consumer stock but tends to trade on the fortunes of its lucrative Cloud business).
Subtract their gains away, and the US market is flat to down in 2023.
Do all these companies deserve to lumped into the AI basket, as investors seem to be doing? I understand the logic – huge data centres, various software opportunities, vast existing customer bases – but count me a sceptic.
Unfortunately for my hairline, I’ve been paying attention through multiple tech revolutions over the past few decades.
And I’d observe that the winners of one are seldom the most dominant companies in the next.
- The video game colossus of the 1970s, Atari, is nowhere today.
- Apple didn’t win the desktop computer war of the 1980s.
- Microsoft was an also-ran for most of the Internet boom in the 1990s – yet many younger Fools probably haven’t heard of the upstarts like AOL and Netscape who outshone it back then.
- Nokia doesn’t make the smartphones we use today. Nor does Blackberry.
The list goes on.
Of course, a few companies do endure or even go on to greater things.
Apple and Microsoft came back bigger and better than ever, eventually, though it was a close-run thing for the iPhone maker. Amazon too.
But they’re the exceptions. It’s as likely the names we will bandy about when we talk about AI in 20 years are today two-person teams renting space from WeWork.
Perhaps the best justification then for investing in the tech behemoths is more the billions that their own venture capital arms have invested into AI start-ups (OpenAI was backed by Microsoft, for example) rather than their existing core operations.
Billions ventured not much gained
The pure-play venture capitalists certainly understand that the future will very likely belong to the new, so they too have poured billions into AI – even in the face of a crash in valuations for other disruptive technology startups.
In fact, 2022 alone saw 3,198 AI startups receive $52.1 billion in funding in 3,396 separate deals, according to the GlobalData Financial Deals Database.
Indeed, I heard one Silicon Valley VC describe this as the best of times if you’re raising money for an AI outfit, but the worst of times for any other kind of firm.
And I find it hard to believe all this money is being invested any more soberly than I think day traders have carefully calculated Nvidia’s likely earnings in a decade.
What’s more, Venture Capital’s love affair with AI goes back to at least 2012, says Deutsche Bank. Its analysts counted more than 175,000 patent entries for artificial intelligence applications made since that year.
Even ‘Modern’ AI – large language models – are a five-old ‘overnight sensation’, yet investors are suddenly as euphoric as if we’d just invented transistors.
Computer says no
Arriving late is a great way to get a quick hit of the buzz of a party in full-swing.
The danger is all the tasty canapés have been eaten, the bubbly was downed hours ago, and you’re left swigging watered-down punch that nobody else wants.
I fear the same could be true of those bidding up the most obvious AI plays today.
At the least it might serve us well to think about investment opportunities further along the AI value chain.
Meta and Alphabet made trillions out of the Internet, not the nuts-and-bolts makers like Qualcomm. You could say the same about Apple and smartphones.
Perhaps the firms that stand to benefit the most from AI are not the ones who create or enable AI technology, but rather the many more that will streamline production and boost profitability through AI’s gradual deployment?
Of course, one way they could do that is by firing – or hiring fewer – expensive, salary-crazy and sleep-demanding humans, by replacing them with AI.
On reflection, maybe the existential questions facing investors and wider society are not so different after all.