IT Essentials: Investors are running out of patience with GenAI

Where are the use cases?

Cash and the planet are being torched in the frenzied pursuit of the AI big win. But is the fire about to be starved of oxygen?

Tech stocks took a hit last week, as Google parent Alphabet warned that whilst it's Q2 revenue growth was broadly in line with expectations, Q3 was less likely to do so. The reason?

"I think we are in this phase where we have to deeply work and make sure on these use cases [for AI products], on these workflows, we are driving deeper progress on unlocking value, which I'm very bullish will happen," CEO Sundar Pichai said.

"But these things take time."

That time took about 5% off Alphabet's value by close of business yesterday.

The story was worse at Tesla, where shares fell more than 12%. Profits have nosedived by 45% year-on-year at the EV giant, and operating margin was reduced to 6.3%. That's despite savage cost cutting and firing 10% of the workforce.

But a key part of the reason that investors are taking flight is another delay in delivering Musk's obsession – the robotaxi. Musk tweeted, quite charmingly earlier this year that "going balls to the wall for autonomy is a blindingly obvious move. Everything else is like variations on a horse carriage".

Problem is, a horse and carriage would have arrived sooner than the robotaxi which has been delayed, delayed and delayed some more. Musk even acknowledged on the earnings call that his predictions on timescales for autonomous vehicles had been ‘overly optimistic,' but provided only wild predictions of Tesla's likely value when and if they finally materialise (we can only hope, safely.)

But Tesla and Alphabet aren't the only companies fuelling the GenAI fire by throwing vast quantities of capital onto it. Microsoft has poured billions of dollars into OpenAI, and in addition to the Anthropic deal, AWS like all its cloud competitors is spending billions on building and buying datacentres for the hardware necessary to run GenAI. Gartner expects hyperscalers to be burning through $180bn a year by 2028.

All the smoke generated by the AI hype fire has obscured the fact that very few organisations have managed to really monetise it as yet. The only company making money out of AI is Nvidia .

Despite all the promises of a revolution in productivity, investors are starting to ask, "where are the use cases?" A realisation is gradually dawning that these improvements could be years away – or further.

The limitations of GenAI as it presently exists such as hallucinations (which somehow sound less damaging than "getting it wrong" or "failing") are better understood. Some of these failings are going to be exacerbated when, as is already likely to be happening, AI models start being trained on data they themselves have generated. Researchers have warned of the prospect of a negative feedback loop leading to "model collapse." The result is a model that cost billions to build eating itself and becoming completely unusable.

A recent survey of people using GenAI tools at work found that 77% said that AI had added to their workload rather than increased their productivity in the way envisioned by their bosses.

GenAI is starting to resemble another, expensive tech bubble. Sure, there's the potential to revolutionise everything but isn't there always? The metaverse, blockchain and the .com boom – the pattern is always the same. Projected dizzying returns, obscene sums of money being hurled at anything with the right badge, then a cratering into the slough of despond (to paraphrase Gartner) as the gap between investment, tangible productivity gains and returns becomes apparent.

So to the billion-dollar question: Are investors running out of patience? The short answer is yes. In addition to the pressure on the share prices of the tech giants, private investment in the form of venture capital and corporate mergers and acquisitions have been tailing down for the last 18 months.

LLMs are evolving too slowly from very expensive novelties to widely adopted and cost-effective tools, and those with the deepest of pockets are beginning to wonder just how long they have to wait to see the returns they were promised by technology with so much potential it could pose an existential threat to humanity itself.

The tech broligarchy are going to need to burn through a lot more cash in the meantime. They just might find it harder to come by than before.