🔒Musk’s xAI: A Billion-Dollar gamble against big tech’s AI supremacy

In a déjà vu move, Elon Musk, co-founder of OpenAI, pledged $1 billion to fund his new AI venture, xAI, challenging Big Tech’s dominance. Musk’s concerns about AI’s impact on humanity resurface as he attempts to surpass OpenAI’s GPT-4. Despite lofty goals, xAI faces an uphill battle, with only $135 million secured from investors. Musk’s unpredictable nature and financial challenges may push him toward partnerships with industry giants, betraying his initial pledge and highlighting the complex dance between ethics and innovation.

Sign up for your early morning brew of the BizNews Insider to keep you up to speed with the content that matters. The newsletter will land in your inbox at 5:30am weekdays. Register here.

Elon Musk’s AI Dreams Are Going in Circles: Parmy Olson

By Parmy Olson


(Bloomberg Opinion) —

 When Elon Musk co-founded OpenAI in late 2015, he pledged to fund the non-profit lab with $1 billion of his own money; he had safety in mind and hoped to counter Google’s stranglehold on AI research. That $1 billion never fully materialized, and Musk left OpenAI on bad terms.

Today he’s trying to back another AI company (his own) with another $1 billion (not his own), according to an SEC filing (although, confusingly, he tweeted on Wednesday they are “not raising money right now”). That not only puts Musk back where he started on AI, it illustrates a pattern among wealthy technologists who’ve been chasing the transformative technology for the past decade: Their worries about how it could hurt humanity tend to get put aside as the race to make their systems more powerful heats up.

It’s been less than a year since Musk signed an open letter calling on developers to take a six-month pause from training AI models. Instead of halting himself, Musk founded xAI, a research lab that competes with OpenAI, Anthropic and Google DeepMind to build systems that could surpass human intelligence, known as artificial general intelligence (AGI).

xAI has a clever logo and an unusual mission statement: to “advance our collective understanding of the universe.” That harks back to another old dalliance of Musk’s. Before backing OpenAI, he also invested in (and eventually exited) the cutting-edge AI lab DeepMind, whose founder Demis Hassabis wanted to build AGI to help unlock the mysteries of reality. 

But the lofty goals of leaders like Hassabis and OpenAI’s Sam Altman often fade to background music as their companies compete to build the most powerful systems and partner with Big Tech firms. OpenAI’s philanthropic goals took a backseat as it became a capped-profit company in 2019 and built technology for its lead investor Microsoft Corp. DeepMind has meanwhile focused on creating Google’s main competitor to ChatGPT, known as Gemini. Even Anthropic, whose founders split from OpenAI over fears it was putting commercial goals over safety, now seems to be aligning themselves with Amazon.com Inc., from which they are taking $4 billion in investment, and with Alphabet Inc.’s Google, which is supplying an additional $2 billion.

Musk, to his credit, doesn’t want to help another tech monopoly consolidate power over artificial intelligence. He is a lone operator, and too capricious for the likes of Microsoft CEO Satya Nadella or Amazon CEO Andy Jassy to handle. Hypocritical as he may be when it comes to “pausing” for AI safety, Musk is at least trying once more to build a counterpoint to Big Tech, something that for many reasons is needed now more than ever.

Unfortunately, this will be an uphill battle for him. Musk has cobbled together $135 million so far from investors, leaving another $865 million to go. Even if he gets the $1 billion he’s after, that might not be enough to cover the expensive, vast computing power he’ll need to create an AI model more sophisticated than the current gold standard, OpenAI’s GPT-4. Microsoft, Google, Amazon and Meta Inc. are the world’s leading makers of large language models for a reason: They alone have the money, talent and computing power to build them well.

For all his noble goals, Musk has tied himself up in knots with this endeavor. Some 25% of xAI’s proceeds will go to investors in X, formerly known as Twitter, but neither X nor xAI look well placed to make a lot of money. Advertisers were fleeing the social media platform even before Musk suggested that they “go f***” themselves, and it’s unclear how he’ll get people to pay for xAI.

Musk’s brand value can only hold out for so long, and he may find himself tempted to partner with a big player like Google or Nvidia. Then he will not only have come full circle, but gone back on one of his biggest pledges for AI. That would hardly be shocking. 

Read also:

© 2023 Bloomberg L.P.

Visited 484 times, 5 visit(s) today