Never Miss a Beat: Get a snapshot of the issues affecting the IT industry straight to your inbox.
Did GenAI Expectations Just Crash Into Cold Economic Reality?
After two years of ChatGPT-fueled GenAI hype, 2024 was supposed to be the year companies would invest and implement emerging AI technology. Does this week’s stock market meltdown mean the AI hype has peaked?
9 Min Read
Andrey Suslov via Alamy Stock
At a Glance
- GenAI faces an uncertain future in the wake of deep stock market losses this week.
- Businesses could be over-inflating GenAI hype with ‘AI washing.’
- GenAI still needs to find its ‘killer app’ for businesses to make money.
Generative artificial intelligence (GenAI) has grabbed daily headlines since OpenAI launched ChatGPT, and the near instant success of the chatbot has fueled a multi-billion-dollar arms race to adopt GenAI.
A Goldman Sachs report says investments in AI will reach $200 billion by the end of the year. According to research firm IDC, enterprises are expected to spend more than $40 billion in GenAI in 2024.
For the businesses producing enterprise AI technology, the increased spending has created a cash cow as companies float resources toward the promised return on investment: AI’s increased productivity and lower costs could create a new tech renaissance.
While COVID-19 produced a temporary tech boom as companies raced to outfit their locked down employees to work remotely, the subsequent years saw decreased profits that led to historic tech sector layoffs. GenAI seemed to be the bright spot in the industry, driving AI chipmaker Nvidia past Apple in market value and spawning hundreds of eager startups.
Just days before the August 5 stock market drop that wiped out $1.3 trillion in market capitalization, Nvidia had announced that its AI chip that would power data centers would be delayed by three months because of a design flaw. The bad news converged with other tech sector woes, including a global IT outage that shut down critical services and took days to remedy. CrowdStrike, the cybersecurity firm at the center of the outage, saw its share prices drop 44% in July.
Big tech firms’ quarterly results were a mixed bag. While Microsoft showed strong revenue growth, it missed revenue projections for its AI endeavors, including the Azure public cloud, Windows Server, developer platform GitHub, and its AI initiative Nuance. Still, those lines of business brought $28.5 billion in revenue for the quarter, a 19% year-over-year increase.
Late to the AI game, Apple won’t know the impact of its newly announced Apple Intelligence features until later this year. The company is pinning hopes on its continued success in iOS devices on the success of its AI features.
Where Is All This Money Going?
IDC identified 42 use cases for GenAI and AI, including AI-enabled customer service and self-service, augmented contact center and field service, augmented connected maintenance, digital commerce, and many other attractive uses. On the vendor side, AI Infrastructure spending is happening, despite economic headwinds. Microsoft, Meta, and Google said they spent a combined $40.5 billion on land, infrastructure, and chips to boost their AI services, according to their most recent financial reports.
Even companies undergoing massive cost cutting efforts aren’t shying away from AI spending. Dell Technologies will slash its workforce (by a reported 10% or 12,500 people this week alone), as it seeks to retool its business to meet AI demand. In an internal memo seen by Bloomberg, Dell said it was creating a “leaner” organization and building “a new group focused on artificial intelligence products and services.” While announcing 15,000 in layoffs, Intel Corp. says it intends to march forward with its ambitious IDM 2.0 chip manufacturing plans.
Data centers will need to keep up with a massive workload demand. In a May report, Goldman Sachs said AI will increase data center power needs alone by 160% by 2030, pushing worldwide AI usage to 3-4% of overall power consumption by the end of the decade. In the US alone, Goldman Sachs predicts up to 8% of our total power use will come from data centers by 2030, compared to 3% in 2022. The US will need to invest about $50 billion in new capacity, while Europe will need $1 trillion to prepare its grid for AI, according to the report.
If the data center market can keep up with that infrastructure growth, it could see exponential revenue growth, growing to $624 billion by 2029, according to Statista. But if the economy slides into recession, that needed investment will be a harder sell, analysts say.
“If there’s an economic slowdown, it’s likely to impact IT budgets overall,” says Arun Chandrasekaran, vice president and analyst at Gartner.
ROI Still Uncertain
Money is certainly changing hands, but what is the return on investment for the average enterprise? There’s little evidence to show that GenAI implementation within an organization is causing a windfall of financial gains. But GenAI providers are reaping the rewards anyway, with many Fortune 500 firms signing on for OpenAI’s Enterprise ChatGPT offerings, and new commercial AI offerings from Salesforce, Oracle, and Cisco. Hyperscalers like Alphabet, Microsoft, Amazon, and Meta are benefiting from AI’s large-scale data and cloud computing needs.
But for the average business, a big AI money-maker remains elusive.
“To some degree this is like the beginning of the Internet,” Paul Dietrich, chief investment strategist at B. Riley Financial, tells InformationWeek. “The AI infrastructure comes first and that requires a lot of capital expenditures … Let’s be honest: The internet was initially funded by the government. AI is being funded be private companies and they know that the dominant AI infrastructure firms will be dominant players forever.”
He adds, “Right now, much of AI is given away for free or almost free. How will companies make money with AI? That will eventually come, but no one knows how money is to be made.”
Gartner’s Chandrasekaran says Monday’s stock drop was not necessarily a “doomsday” event for GenAI, but instead offers enterprises a chance to “recalibrate” and look to establish better data governance and use cases before making large investments.
“There are clearly a lot of folks in the media pressing the ‘doomsday’ button and a lot of people still think [GenAI] is transformational,” he says. “I think the truth is somewhere in between. The underlying technologies are still developing quite rapidly. What’s changed is enterprises, broadly speaking, are struggling to get value from GenAI because their own data governance foundations are not very strong.”
And companies are still looking for a pay day. “At the end of the day, investors are expecting a return,” Chandrasekaran says. “The fundamentals of how businesses are valued and the expectations from investors … that won’t change.”
In a note following Monday’s stock market selloff, Dietrich said his previous warnings about a potential recession and market crash could materialize if AI enthusiasm begins to wane. “Eventually, we will have another long-term bear market recession,” he wrote.
Monetization at the enterprise level might hinge on differences in generational use.
A new study by Zuora shows a gap between GenAI enthusiasm and actual use, noting that 72% of consumers surveyed are currently not using GenAI services. Younger people are more likely to use GenAI services. But 70% of survey respondents said they were not willing to pay extra for embedded, GenAI-powered features.
“The varying interest between generations highlights the need to deliver clear value while acknowledging that value may differ for each age group,” Amy Konary, senior vice president at Zuora, tells InformationWeek.
Dirty Laundry: The ‘AI Washing’ Problem
The public was enamored with GenAI after seeing the power of ChatGPT unleashed in 2022. The company reached 100 million users within weeks. Businesses followed and excitement reached a fever pitch in 2023.
That enthusiasm sparked massive demand for GenAI implementation. Investors and boards were demanding immediate investment in AI, and specifically, GenAI.
Some companies began attaching the terms “intelligent,” “AI”, or “machine learning” to products and services that had at best a vague or tangential relationship to AI, and at worst no connection to large language models or machine learning.
The name for this dubious practice is “AI washing,” modeled after the better-known term, “greenwashing,” which involves a company or institution making misleading claims about environmental and sustainability efforts. Analysts worry this practice adds to inflated hype and could fuel investor fears that GenAI lacks a tangible business use advantage.
Some big corporations that have been accused of AI washing in the last year include Coca-Cola, Pepsi, Zara, H&M, McDonald’s, and Philip Morris. The US Securities and Exchange Commission (SEC) has promised a crackdown on AI washing.
During a speech at an NYC event in April, Gurbir S. Grewal, director of the SEC’s enforcement division, said, “While perhaps not quite yet a perfect storm, there’s certainly one brewing around AI, and it is incumbent on each of us to make sure it does not come to pass and that investors are not harmed by noncompliance with securities laws when it comes to this new technology.”
Grewal said the SEC would use its experience enforcing laws against greenwashing in future AI washing efforts. In September 2023, Deutsche Bank subsidiary DWS Investment Management Americas Inc. agreed to pay $19 million to settle the SEC’s charges of greenwashing.
“I hope these actions put the investment industry on notice. If you are rushing to make claims about using AI in your investment processes to capitalize on growing investor interest, stop,” Grewal said.
Zoura's Konary says AI washing won't stand up to increasing customer understanding of the technology. "The findings are clear," she says. "AI washing will not cut it when looking to drive revenue related to the AI hype. Customers are getting smarter, faster, and expectations are rising quickly. Companies need to ensure a clear high return on investment and that AI solutions deliver benefits and value."
Will AI Have an iPhone Moment?
CIOs are now likely going to face tightening budgets and need to re-evaluate their GenAI implementation strategies.
“I think from an enterprise standpoint, if I’m a CIO, the way I’m probably reading the tea leaves is we need to recalibrate our expectations,” says Gartner’s Chandrasekaran. “We need to get back to basics and get our data foundation, our governance foundation stronger, and we need to set more realistic expectations around business value and ROI that’s going to come.”
Finding value will likely take another moment like OpenAI’s ChatGPT launch -- an application that will help enterprises realize enough revenue to make the endeavor worth the risk.
"Organizations are prioritizing AI innovation, but to stay viable, they will need to understand how to effectively monetize," Zuora's Konary says. But ROI will take time, she says. "We are in the early days, and ROI will remain elusive for both customers and vendors while we are in the experimentation phase. Chasing that metric requires an understanding of the value proposition, including willingness to pay, and behavioral changes."
B. Riley’s Dietrich tempers his bearish economic outlook with a little hope for the future of GenAI:
“Someone will do for AI what Apple did for the internet with the iPhone. No one foresaw the Apple iPhone when the internet was in its early days. Out of iPhone came Facebook, Amazon, and all apps,” he says.
About the Author
You May Also Like
More Insights
