Feature
Believe it or not, not everyone is convinced that AI is the future.
A small group of analysts and financial experts say the excitement is generated, at least in part, by overly optimistic forecasts and best-case scenarios, which ignore the challenges that stand between the real world and AI’s promise.
The Argument Against AI
Some of the arguments are rooted in the behavior of AI companies themselves. As they steadily launch new and improved versions of their software, AI-naysayers say these companies gloss over critiques, talk up their products’ potential and brush aside talks of AI’s impending state of consciousness.
The media's cheerleader-style coverage isn't helping either, critics say. OpenAI’s employees often treat ChatGPT as if it was ready to pull up a chair and chat, beginning their day by asking, “Hey, ChatGPT, how’s it going?” Kevin Roose writing for The New York Times asked, “[If] OpenAI’s own employees can’t resist treating ChatGPT as a human, is it any mystery whether the rest of us will?”
The answer isn’t as black and white as the Times would have you think. Few people have as much skin in the AI game as OpenAI’s professionals do, and it’s hard to imagine anyone who might be more invested in the idea of AI’s intelligence growing by leaps and bounds. As Bloomberg's Max Chafkin notes, a new version of software doesn’t always represent obvious progress: The newest version of ChatGPT, for instance, brings with it many of the same drawbacks as previous iterations.
Meanwhile, developers of other large language models continue to talk a better game than they play. If you’re looking for instructions on making pizza, Google’s AI-driven search results function can serve up instructions — including the suggestion to add glue to your recipe to hold the cheese in place.
AI proponents argue that such glitches come with the territory of advanced technology development, and developers maintain that nonsensical responses such as putting glue on pizza are rare — or the result of unusual, rarely posed questions. “The examples we've seen are generally very uncommon queries and aren't representative of most people’s experiences,” a Google spokesperson explained to Sky News.
Related Article: Are We Focusing on the Wrong AI Use Cases?
The (Long) Road Ahead to GenAI Maturity
Beyond the glitches, which many skeptics say prove generative AI has a long way to go, the argument against AI for now is economics.
Jim Covello, head of stock research at Goldman Sachs, said in an interview with The New York Times that the costs of implementing AI are too high and the benefits too sparse to justify both the hype and the spending we see today.
By some estimates, the Times reported, to be widely implemented AI could require an investment of $1 trillion in data centers, utilities and applications. Covello said that’s too much for a technology that’s prone to make mistakes.
Tech budgets are already feeling the impact of AI’s demands. According to Gartner, spending on data centers will increase by 10% during 2024, compared to 4% in 2023. John-David Lovelock, distinguished vice president analyst at Gartner, sees “gold rush level spending by service providers in markets supporting large scale GenAI projects, such as servers and semiconductors … In 2024, AI servers will account for close to 60% of [large data centers’] total server spending.”
Overall, spending on AI applications, infrastructure and related IT and business services will more than double by 2028 to $632 billion, estimates IDC. Between 2024 and 2028, the industry will see a CAGR of 29%, largely due to the introduction of AI into a range of products, like smartphones or business-software packages.
Exactly how much revenue such investments will generate is unclear. While OpenAI’s annualized revenue has reached some $3.4 billion, relatively few AI companies have been able to push their revenue into the $100 million range. That’s partly because the real demand for AI is unclear, said Sequoia’s David Cahn.
“Outside of ChatGPT, how many AI products are consumers really using today?” he asks in a company post. “Long term, AI companies will need to deliver significant value for consumers to continue opening their wallets.”
Related Article: AI Overload: Is the Hype Losing Steam?
A Contrarian Voice in the AI Hype Cycle
The arguments are sound, but many disagree with Covello and others in their case against AI — even inside Goldman. George Lee, co-head of the firm’s geopolitical advisory business, argues — like most AI proponents — that the technology will save workers time and increase productivity.
“The long-term impact of platform shifts is that applications emerge over time as that technology is refined, made more readily available, made cheaper,” he told the Times in the article mentioned earlier.
Covello’s not convinced. In the past, costs have risen, not fallen, for many sophisticated technologies, he argues, like those used to make semiconductors. And, he wonders, is AI fulfilling a pressing need? “Overbuilding things the world doesn’t have use for, or is not ready for, typically ends badly,” he said.
Conference
Sep
16
HR Tech Conference Las Vegas 2025
Register
Conference
Oct
27
Gartner HR Symposium/Xpo Orlando 2025
Register
Covello is also unmoved by comparisons of the AI rush to the World Wide Web’s introduction in the early 1990s, noting the economic troubles caused by the dot-com boom — including the loss of millions of jobs. That’s “not what anybody should be rooting for.”
