Arm plans an IPO as AI fever sweeps Wall Street

 

By Mark Sullivan

 

What the expected Arm IPO could mean for AI

The semiconductor giant Arm is getting ready to hold a public offering in what will likely be 2023’s biggest IPO. In a new filing, the company says it’s looking to raise $4.87 billion from selling just 9.8% of its shares on the Nasdaq exchange. If all goes well, Arm could end up with a valuation of $52 billion. Arm was a public company until 2016 when SoftBank took it off the market for $32 billion. 

Much of the excitement around the IPO is due to the market’s current interest in AI. Many investors are looking for investment opportunities with companies (other than Nvidia) that could directly profit from the AI boom. Arm likely won’t pose a threat to Nvidia’s dominance providing the A100 and H100 chips that are widely used within data centers to train large AI models. Rather, Arm’s AI play will be in designing chips that run smaller or compressed versions of AI models within devices (sensor modules or smart speakers, perhaps), Pitchbook emerging tech analyst Brendan Burke tells me. But institutional investors don’t seem convinced that the AI boom will quickly expand the edge chip market. Softbank, in fact, adjusted down a $64 billion valuation of Arm from earlier this year, Burke points out. “The lowered price range relative to the initial target shows that investors are exercising caution in assigning expectations for AI growth,” he says.

 

Survey: University AI experts favor a federal AI agency

University computer science professors and researchers aren’t thinking as much about AI wiping out mankind as they are about getting a regulatory grip on the booming technology. A new survey conducted by Axios and Syracuse University asked 213 experts from 65 top computer science programs what entity would be best for regulating AI. The results show that 37% favor the establishment of a new government agency to oversee and create guardrails around tech companies developing potentially harmful models. Another 22% said a global organization or treaty is needed. And while 16% said Congress could be relied upon to pass effective regulation, only 3% believed the tech industry could regulate itself. 

The survey results come days before Senate Majority Leader Chuck Schumer is set to host a slew of tech leaders and experts at an AI forum on September 13. The guest list is a who’s who of industry titans: OpenAI CEO Sam Altman, Alphabet CEO Sundar Pichai, Meta CEO Mark Zuckerberg, Microsoft CEO Satya Nadella, X CEO Elon Musk, Inflection AI founder Mustafa Suleyman, former Google CEO Eric Schmidt, and a host of others. All will be traveling to D.C. to discuss approaches to regulating AI.

 

Fast Company

(10)