Is reality crashing down?
It's a gold rush in the artificial intelligence sector, as companies large and small rush to stuff buzzy machine learning tech into their products, from Expedia's travel chabot to BuzzFeed's AI-powered quizzes.
Whether it'll pay off in the long run is anyone's guess. But not all experts are so sure, with one analyst telling CNBC that he thinks generative AI is going to run into some major reality problems that amount to a "cold shower" by next year.
"The bottom line is, right now, everyone’s talking generative AI, Google, Amazon, Qualcomm, Meta," CCS Insight chief analyst Ben Wood told the broadcaster. "But the hype around generative AI in 2023 has just been so immense, that we think it’s overhyped, and there’s lots of obstacles that need to get through to bring it to market."
One key issue: any company that wants to develop its own AI from scratch — as opposed to licensing it from the likes of OpenAI — is going to need expensive computer chips from the likes of Nvidia and capital to service them.
"Just the cost of deploying and sustaining generative AI is immense," said Wood. "And it’s all very well for these massive companies to be doing it. But for many organizations, many developers, it’s just going to become too expensive."
In the interview, Wood also anticipated other factors leading to an AI slowdown, including governments around the world enacting new regulation of the tech.
Adding to those woes, public hype may be starting to fade as users realize the tech's penchant for making up facts or offering glib, nonsensical answers to queries. And maybe the biggest question: how are AI startups planning to actually make money?
The Wall Street Journal reported yesterday even major players like Microsoft, Google and Adobe are struggling to figure out how to turn a profit from AI, since the tech is so expensive to run and the market is flooded by free offerings. Even worse, the costs jack up as more customers use AI products.
As an example, WSJ reported that Microsoft is losing money with its popular GitHub Copilot, which it launched with the Microsoft-backed OpenAI.
Some companies think AI will get less expensive with time like previous technologies, according to the WSJ, but venture capitalists are still becoming wary around AI hype.
Will the tech be around long term, at least in some form? Probably. But it could be that the long road to a sustainable business model is much longer and more winding than its early advocates thought.
More artificial intelligence: Study Finds That Police “Crime Predicting” AI Fails Miserably at Predicting Crimes
Share This Article