Worries about AI’s risks to humanity loom over the trial pitting Musk against OpenAI

OAKLAND, Calif. — On the coronary heart of the trial pitting Elon Musk towards OpenAI CEO Sam Altman is a second once they discovered frequent trigger on an ever extra urgent query: how one can shield humanity from the dangers of artificial intelligence.

It turned bitter, and the jury is charged with settling the following authorized dispute between the 2 Silicon Valley titans.

However the unresolved questions concerning the risks of AI have been looming over the federal courthouse in Oakland, California, for the reason that trial started final week. The expertise itself shouldn’t be on trial – the choose has warned attorneys to not get “sidetracked” by questions on its risks – however witness testimony has touched on issues round workforce disruptions and the prospect raised by Musk that superhuman AI may someday kill us all.

Musk, the world’s richest particular person, filed the case accusing his fellow OpenAI co-founder of betraying guarantees to maintain the corporate as a nonprofit. Altman, in flip, accuses Musk of attempting to hobble the ChatGPT maker for the good thing about his personal AI firm.

One witness, AI pioneer Stuart Russell, mentioned that the “winner take all” energy wrestle over AI’s future is itself threatening humanity.

Musk’s attorneys introduced Russell to the stand as an skilled witness, on the fee of $5,000 an hour. The College of California, Berkeley laptop scientist listed a number of AI risks, from racial and gender discrimination to jobs displacement, misinformation and emotional attachments that take some AI chatbot users down a spiral of psychosis.

“Whichever firm develops AGI first would have a really massive benefit” and an more and more massive lead over everybody else, Russell advised the court docket, utilizing the initials for artificial general intelligence, a time period for superior AI expertise that surpasses people at many duties.

The trial facilities on the 2015 delivery of OpenAI as a nonprofit startup primarily funded by Musk.

Each Musk and Altman, who has not but testified within the trial, have mentioned they needed OpenAI to securely develop AGI for the good thing about humanity and never for anybody particular person’s acquire or underneath anybody particular person’s management. And each camps allege it’s the opposite man who was attempting to regulate it.

A jury of 9 folks chosen from the San Francisco Bay Space will get to say which one in every of them is telling the reality.

Early on, Choose Yvonne Gonzalez Rogers warned attorneys, notably Musk’s, to not delve into broader AI issues that transcend Musk’s claims that OpenAI violated its charitable mission.

“This isn’t a trial on the protection dangers of synthetic intelligence. This isn’t a trial on whether or not or not AI has broken humanity,” Gonzalez Rogers advised attorneys earlier than jurors arrived on the federal courthouse.

Nonetheless, Musk managed to skirt that steering in his testimony final week. Requested to explain synthetic normal intelligence, Musk mentioned it’s when AI turns into “as good as any human,” and added that “we’re getting near that time,” and AI might be smarter than any human as quickly as subsequent 12 months.

Musk mentioned he has “excessive issues” about AI and has had them for a very long time. Musk mentioned he needed a “counterpoint” to Google, which on the time had “all the cash, all of the computer systems and all of the expertise” for AI, with no counterbalance.

“I used to be involved AI could be a double-edged sword,” he mentioned.

Throughout his testimony, Musk repeatedly mentioned that he might have based OpenAI as a for-profit firm, similar to the opposite corporations he began or took over. “I intentionally selected this,” he mentioned, “for the general public good.”

The choose expressed some skepticism. In feedback to attorneys final week earlier than the jury got here into the room, Gonzalez Rogers identified that Musk, “regardless of these dangers, is creating an organization that’s in the very same area,” referring to the billionaire’s xAI synthetic intelligence firm, which launched in 2023 and has since merged with Musk’s rocket firm SpaceX.

OpenAI’s facet additionally claims its objectives are to profit the general public. OpenAI co-founder and president Greg Brockman, a defendant in Musk’s lawsuit together with Altman and their firm, mentioned he thought the expertise OpenAI was growing was “transformative” — greater than companies, company buildings and larger than anybody particular person. It was, he mentioned, “about humanity as a complete.”

Brockman testified this week that his No. 1 objective was at all times the “mission” of OpenAI and it was Musk who sought unilateral management over the corporate.

Brockman recalled a gathering the place at first Musk appeared open to the thought of Altman being OpenAI’s CEO. In the long run, nonetheless, “he mentioned folks wanted to know he was in cost.”

Along with damages, Musk is searching for Altman’s ouster from OpenAI’s board. If Musk wins, it might derail OpenAI’s plans for an preliminary public providing of its shares.

___

O’Brien reported from Windfall, Rhode Island.