Linus Torvalds, creator of the Linux kernel, believes that much of the industry’s marketing about generative AI is just flair with no real substance – and it may take many years before the technology is proven.
The reformed pottymouth was speaking at the Open Source Summit in Vienna last month about a “video-centric storytelling platform.” TFiR when asked for his thoughts on modern technologies, particularly GenAI.
“I think AI is really interesting and I think it’s going to change the world and at the same time I hate the hype cycle so much that I really don’t want to go there, so my approach to AI in this moment is: basically ignore it,” Torvalds said.
“I think the whole tech industry around AI is in a really bad position and it’s 90 percent marketing and 10 percent reality. In five years, things are going to change and at that point we will see how much of the AI is used for real-world workloads,” he added.
OpenAI’s ChatGPT “does great demos and is obviously used in many areas, for graphic design and things like that, but I really hate the hype cycle.”
The IT industry is notorious for its marketing bluster, focusing on nascent technology only to overpromise and underdeliver. This, the Linux kernel developer said, is a problem: “Before AI, a few years ago, the only thing people talked about was cryptography and I just don’t like the hype cycle” .
So far, some $95 billion has been invested in AI startups since the start of 2023 and the big four hyperscalers have spent $200 billion in capital expenditures, according to Steve Brazier, a member of ‘Informa, who spoke this month at the Canalys Channels Forum.
“With about $200 billion in investment, only about $20 billion in revenue actually comes from consumers and businesses in terms of AI services, like Copilot licenses and ChatGPT licenses, which is a very poor return in terms of actual results in terms of end users. And the whole gamble (whether the AI explosion continues or not) will depend on their ability to raise that $20 billion as quickly as they hope. »
The investor spotlight will once again be on Google later today and Microsoft tomorrow when these companies report their latest financial results. Microsoft has invested billions in OpenAI, Mistral and Inflection, while Google has bet on Anthropic.
Both companies are still trying to convince customers of the productivity benefits of GenAIalthough the the return on investment seems uncertainwith unforeseen corporate governance challenges sometimes being the sticking point.
Gary Marcus, professor emeritus at New York University and AI expert, said The Reg last week that GenAI has its uses but “is not that reliable.”
“More broadly, everyone is pushing GenAI to try to make its huge investments profitable, but it’s not going very well. In 2023, there was nothing but hype; in 2024, I see a lot of disillusionment,” Marcus said.
So it seems that Torvalds is not alone. He relies on money and – unlike chatbots that go off the rails – he didn’t even need to say a simple rudeness to take stock.
