Amazon Web Services boss calls AI business ‘massively overrated’

Adam Selipsky, CEO of Amazon's dominant cloud computing platform AWS, sees parallels between the buzz around AI and the rush of companies hoping to capitalize on the emerging technology to the dot-com bubble of the 1990s.

“If you go back to 1997 and ask, ‘Is the Internet overrated or underrated?' I argue that, as he told Wired at a Harvard Business School conference over the weekend: “But then, ‘Were the leading companies at the time incredibly overvalued?' If you ask, yes they were.”

He later returned to his broad dismissal, explaining that “some AI companies at the center of the storm have faltered significantly” and that the future of companies launching MVPs to demonstrate AI methods may collide with cold, hard reality.

“Once you get into production, it can be very expensive,” he said.

Selipski is not alone in his cautious view. OpenAI is the company that everyone thinks about when the word AI is mentioned, but its COO Brad Lightcap said in a recent interview that the notion that AI “can make a tangible business difference” is overblown.

“There's nothing you can do with AI that will completely solve this problem,” he concluded, and pitched himself to those who are more into open-source AI than Emad Moustaku—CEO of Stability AI, developers of Stability Distribution. In the same camp as Selipski and Lightcap. He thinks that AI can cause what he calls “the biggest bubble of all time”. oh

Bubble or not, Amazon stands to make a lot of money from a growing industry.

As one of the world's largest cloud computing service providers, AWS is a critical platform for AI developers who need massive amounts of computing power to build and run complex models. Regardless of the hype, Amazon is making big money selling the shovels that will keep everyone alive during the AI ​​gold rush.

Last week, Amazon released its quarterly financial results. AI was a strong part of the mix.

The company described a new expansion of its relationship with Salesforce, which will use AWS to “deepen the data on AI integrations” between the two companies. Healthcare industry stalwart Merck, e-commerce retailer The Very Group, Japan-based Mitsubishi, Korea-based LG, and several other companies have been touted for their AI initiatives.

The computer infrastructure business is now effectively an oligopoly. Amazon, Microsoft, and Alphabet (Google) are the go-to sources for AI developers looking for cloud hyperscalers to train their models. They dominate the market.

But AWS isn't just selling infrastructure, it's also working on its own AI products.

The tech giant recently announced a $4 billion investment in AI startup Anthropic, creator of Chatbot Cloud. It is one of the biggest rivals to challenge OpenAI's ChatGPT. The company is developing its own AI assistant: Q So far, its feature list includes an image generator and a proprietary large-scale language model (LLM) that adds Alexa interactions based on user queries.

“Innovative Genie capabilities like Bedrock, Q and Trainium have connected with customers and are beginning to be reflected in our overall results,” Amazon CEO Andy Jassy said in the report.

Selipsky's caution is coupled with an awareness of AI's potential. “We believe Generative AI will make a difference,” he said. And independent analysts agree. According to estimates by experts including McKinsey and JP Morgan, AI business could add more than $4 trillion to the global economy.

With AIA, as with any gold rush, selling shovels can be more profitable than chasing speculative claims. For Amazon, though, digging some holes on their own doesn't seem like a bad idea.

Edited by Stacy Elliott.