- Elon Musk is suing OpenAI over concerns that it's no longer fulfilling its original mission.
- It's a legal battle that has started to split Silicon Valley on moral grounds.
Elon Musk's lawsuit against Sam Altman's OpenAI hasn't just triggered a battle between two former cofounders of a nonprofit now being derided as a "de-facto subsidiary" of Microsoft.
It has reignited the savior complex of Silicon Valley's biggest personalities too.
At the heart of Musk's lawsuit against OpenAI, which was filed on March 1, is the view that the company has strayed far from its founding mission to create artificial general intelligence (AGI) that benefits humanity.
Musk's lawyers argued the company's new board — formed in November in the aftermath of an attempted coup of Altman's reign as CEO — had resorted not just to developing but to "refining an AGI to maximize profits" for its multibillion-dollar backer Microsoft.
Though some questions are being asked about how much of a case Musk has here, the whole saga has offered plenty of incentive to Silicon Valley's more opinionated leaders to moralize on the future of AI.
Which way, AI man?
Leaders spearheading the rollout of AI have had an important question to ask themselves before releasing their technology to the world: open or closed?
The choice — a crux of Musk's argument against OpenAI — is critical as it presents two vastly different alternatives for letting AI loose.
The open option — in theory at least — champions AI models that are transparent about how they're trained and collectively developed by a global community of developers. Meta has committed to this approach with Llama 2; so too has French startup Mistral AI.
Though concerns exist about open source models being vulnerable to abuse by bad actors, advocates tout its benefits as a huge plus over closed models such as OpenAI's GPT-4, which don't share the data they use for training behind closed doors.
So much so, that they're getting righteous about it.
Some of this moralizing came into the fore over the weekend after venture capitalist Marc Andreessen decided to respond to contentions with Musk's legal tussle with OpenAI from Vinod Khosla, who bet $50 million on the company in 2019.
By Khosla's measure, OpenAI's approach needs to be weighed against a backdrop of national security. "We are in a tech economic war with China and AI that is a must win," he wrote on X, before asking Andreessen if he would open-source The Manhattan Project.
Andreessen ran with the analogy that likened today's AI to nuclear weapons production during World War II.
He used it to explain how, if AI really was as important a technology to keep under wraps as military weapons, it shouldn't be left to a few folks in San Francisco to shield from, say, an espionage campaign led by the Chinese Communist Party.
"What you'd expect to see is a rigorous security vetting and clearance process for everyone from the CEO to the cook, with monthly polygraphs and constant internal surveillance," Andreessen wrote on X, before pointing out that clearly isn't what is seen at OpenAI.
Andreessen acknowledged that the analogy was a bit "absurd," what with AI being math, not nukes.
Of course every part of this is absurd.
— Marc Andreessen (@pmarca) March 3, 2024
(1) AI isn't nukes, it's math.
(2) Big companies, small companies, independent developers, and open source communities should be free to pursue AI.
(3) AI companies should not be nationalized, de jure or de facto (government-protected… https://t.co/2VqL6T5qZe
But his verbal sparring with Khosla online goes a long way towards showing how the case for and against closed-source models like GPT-4 is putting Silicon Valley's moral values in the spotlight.
How the actual legal battle plays out will be closely watched. OpenAI told employees that it "categorically disagrees" with the lawsuit in an internal memo sent to employees, according to Bloomberg, and reckons Musk regrets no longer being involved.
But which side of the battle the rest of Silicon Valley's AI leaders fall on will be watched just as closely. After all, their values could be up for judgment.