Unlock Editor's Digest free of charge
FT editor Roula Khalaf selects her favourite stories on this weekly newsletter.
There are several truisms relating to tech bubbles. One of them is that it's often difficult to identify a bubble while you're in the midst of one: Every single spending or investment decision can seem rational, even when the online effect seems extreme.
When there’s widespread agreement that a bubble is forming, it may well often grow much further, and investors who get out early miss out. And after a bubble bursts, it may well take years to work out whether it was just the product of hype or the harbinger of a fair greater tech boom.
As US technology firms head into the newest earnings season, a certain giddiness is spreading on Wall Street. The middle of 2024 must be something of a “respiratory hole” for generative artificial intelligence. The investment boom triggered by the technology is all too evident, but it’s going to take time for all this latest capability to be put to productive use by the technology industry's end users. Wall Street's patience will soon be tested.
Software firms best positioned to capture the worth of latest technology by embedding it as a feature of their existing products are struggling amid the gloomy stock market backdrop.
Nor has AI produced compelling latest consumer services, despite Apple's recent promise to liberally distribute the technology across its devices. For many individuals, the invention of ChatGPT sparked a wave of pleasure. But unlike the primary time they picked up an iPhone, used Google's search box, or found their friends on Facebook, it hasn't modified their digital lives.
At best, this implies a delay within the widespread adoption of generative AI. At worst, it shows that the technology isn’t as transformative because the tech firms claim. The longer the pause lasts, the more stark the gap between the investment boom and weak final demand becomes.
This is made much more clear by the undeniable fact that this investment boom is resting on a really narrow base. Nvidia revealed late last yr that greater than half of its data center chip sales got here from cloud firms – a market dominated by a handful of huge players. Any indication from the massive tech firms during this earnings season that they’re curbing spending could be a serious blow.
But as tech firms prepare to announce their latest results, all signs point to the boom still being in full swing. Many enterprise customers have only just begun their first pilots of the technology and can increase their testing of the technology in the approaching months, even when it continues to be unclear what ultimate use they are going to find for it. It has also change into a strategic imperative for the massive tech firms themselves to pour money into large language models and the infrastructure that supports them. If machines that may “understand” language and pictures represent a completely latest computing platform, as many within the tech world expect, then all of the foremost players must get a stronger foothold within the technology.
It's also price noting that these firms have good enough financial firepower to sustain and even escalate the fight. The combined operating money flow of Apple, Microsoft, Alphabet, Amazon and Meta increased 99 percent over the past five years, reaching $456 billion in 2023. That was good enough to handle capital expenditures that bloated 96 percent to $151 billion.
Meanwhile, chipmaker Nvidia's next major product cycle, based on its latest Blackwell chip architecture, isn’t scheduled to start until the second half of the yr. The associated lower costs of coaching and running large AI models have guaranteed a rush of shoppers, although demand for the sooner generation of chips stays strong.
There's a paradox here that's common to all latest technology platforms. If the price of using the brand new technology falls, customers could theoretically get away with buying less. But rapidly falling costs normally result in latest uses being found, and demand skyrockets as an alternative. As with the whole lot related to generative AI, this is occurring at warp speed, and it's hard to say how much this may resemble other disruptive technology cycles.
At some point, in fact, all of those investments should yield a return. If not, then the CEOs who’ve been pressured by their boards and fear of competition to require their firms to make use of generative AI will eventually lose patience and move on to other tasks. But all indications are that we are usually not there yet.
