While the hype continues to grow behind enterprise AI deployments and research finds surprising levels of consumer support for the technology, one analyst company believes that the bubble is about to burst.
“Unrealistic expectations that cannot be met and excessive investments that cannot possibly be paid back” is the stark message about artificial intelligence this week from Riot Research, part of Rethink Technology Research.
The company’s latest report finds that the AI market will “only” reach $39 billion globally by the end of 2023. In some sectors, AI will “flop dismally” over the forecast period, while it will thrive in others, it says.
A cull of VC-backed startups in the AI marketplace is expected during 2019 and 2020, continues the report. In the aftermath, it says, there will be “a clear pattern of AI survivors in key vertical and horizontal sectors.
“This is a predictable repeat of earlier infamous bubble-burstings, such as the dotcom collapse of 2001, which lost investors billions of dollars, but from whose ashes great tech giants such as Google and Amazon emerged.”
However, this misrepresents the facts: Google and Amazon were formed in 1998 and 1994, respectively.
Sector hits and misses
The report, AI: Show me the Money, identifies the industrial sectors in which AI will have the biggest impact.
Verticals where AI is set to thrive over the next five years include cybersecurity, automotive, healthcare, manufacturing, and finance and insurance, with the overall verticals market reaching a global value of $24.8 billion by 2023.
Two major horizontals – machine vision and natural language processing – will hit $14.1 billion and $15.0 billion, respectively. But about half of the $29 billion horizontal market overlaps with the vertical sectors, because both machine vision and natural language processing are part of wider AI offerings, says the report.
So why will the bubble burst?
Rethink Research fellow and report author, Philip Hunter, said, “Huge sums have been invested in AI, with the upward trend still accelerating through 2018. Global venture-capital-based investments alone have risen from $3.2 billion in 2014 to $12 billion in 2017, while the number of funding rounds per year for AI startups doubled to around 1,300 over that period.
“The total invested globally in AI during 2018 alone amount to over $100 billion, taking account of money spent by governments and big corporations, as well as VC-funded startups.”
That is more than double the expected annual return from AI even by 2023, so there is no way that many investors will see any payback over the forecast period, explained Hunter. As a result, investors will begin “rolling up startups which fail to generate revenues into others which show promise during 2019”.
The only way that most AI startups have made money so far is from being acquired, adds the report, rather than selling viable products or services.
“Valuations have been based purely on the assessment of the people working for the company, often at as much as $10 million a head. This dangerous method of valuing startups is insane – key people can leave after an acquisition, unless golden handcuffs tie them to the deal, and even that can lead to disenchantment on both sides.
“After the bubble bursts, more realistic and sustainable valuations will be placed both on AI companies and their engineers.”
Internet of Business says
All peaks are accompanied by troughs – the characteristic pattern of waves – and the AI sector is unlikely to be any different.
The millennial dotcom bust was caused by many things: startup employees finding that their stock options were worthless and they couldn’t live on hope alone; businesses based on clicks and click-throughs rather than revenues; overvalued IPOs; absent business plans; trendy offices with no customers, and so on.
Plus a significant flattening of the PC market was taking place in the background, just as the mobile market is levelling off today.
However, a key difference now is that Amazon, Google, Apple, Oracle, SAP, Salesforce, Microsoft, Dell, Cisco, Facebook, IBM, and the rest, all have deep pockets and are snapping up AI startups, and – as the report suggests – their talented leaders or employees. Selling out to a major is the endgame for many startups.
But that’s not to say that this analyst report is wrong. Set alongside the Capgemini report, published yesterday, the precise location of the crunch point should be becoming clear.
While consumers see the technology as being about improved customer service and friction avoidance, most enterprise users see things differently; for them, AI is primarily about slashing costs and seeing swift returns – the recurring strategic error in all new technology deployments.
Given that many organisations are, according to Capgemini, misapplying the technologies for short-term tactical gain rather than long-term strategic advantage, this increases the risk that they may never find the ROI they’re looking for in simple terms. That’s where the market correction will kick in.
In the meantime, hundreds of startups will fall by the wayside, as they have always done, and a measure of realism will emerge. In short, after the peak comes the trough, but the direction of the wave remains clear.