Musk bashes OpenAI in deposition, saying ‘nobody committed suicide because of Grok’

In a recently released deposition filed successful Elon Musk’s lawsuit against OpenAI, the tech enforcement attacked OpenAI’s information record, claiming that his company, xAI, amended prioritizes safety. He went truthful acold arsenic to accidental that “Nobody has committed termination due to the fact that of Grok, but seemingly they person due to the fact that of ChatGPT.”

The remark came up successful a enactment of questioning astir a public letter Musk signed successful March 2023. In it, helium called connected AI labs to intermission improvement of AI systems much almighty than GPT-4, OpenAI’s flagship exemplary astatine the time, for astatine slightest six months. The letter, which was signed by implicit 1,100 people, including galore AI experts, stated determination was not capable readying and absorption taking spot astatine AI labs, arsenic they were locked successful an “out-of-control contention to make and deploy ever much almighty integer minds that nary 1 – not adjacent their creators – tin understand, predict, oregon reliably control.”

Those fears person since gained credibility. OpenAI present faces a series of lawsuits alleging that ChatGPT’s manipulative speech tactics person led respective radical to acquisition antagonistic intelligence wellness effects, with immoderate dying by suicide. Musk’s remark suggests that these incidents could beryllium utilized arsenic fodder successful his lawsuit against OpenAI.

The transcript of Musk’s video testimony, which took spot backmost successful September, was filed publically this week, up of the expected assemblage proceedings adjacent month.

The lawsuit against OpenAI centers connected the company’s displacement from a nonprofit AI probe laboratory to a for-profit company, which Musk claims violated its founding agreements. As portion of his arguments, Musk claims that AI information could beryllium compromised by OpenAI’s commercialized relationships, arsenic specified relationships would spot speed, scale, and gross supra information concerns.

Since that recording, xAI has faced information concerns of its own, however. Last month, Musk’s societal web X was flooded with non-consensual nude images generated by xAI’s Grok, immoderate of which were said to beryllium of minors. This led the California Attorney General’s bureau to open an investigation into the matter. The EU is besides running its ain investigation, and different governments person taken action, too, with immoderate imposing blocks and bans.

In the recently filed deposition, Musk claimed helium had signed the AI information missive due to the fact that “it seemed similar a bully idea,” not due to the fact that helium had conscionable incorporated an AI institution looking to vie with OpenAI.

“I signed it, arsenic galore radical did, to impulse caution with AI development,” Musk said. “I conscionable wanted to — AI information to beryllium prioritized.” …

Image Credits:imgflip

Musk besides responded to different questions successful the deposition, including those astir artificial wide intelligence, oregon AGI — the conception of AI that tin lucifer oregon surpass quality reasoning crossed a wide scope of tasks — saying “it has a risk.” He besides confirmed that helium “was mistaken” astir his supposed $100 cardinal donation to OpenAI; the second amended complaint successful the lawsuit puts the existent fig person to $44.8 million.

He besides recalled wherefore OpenAI was founded, which, from his perspective, was due to the fact that helium was “increasingly acrophobic astir the information of Google being a monopoly successful AI,” adding that his conversations with Google co-founder Larry Page were “alarming, successful that helium did not look to beryllium taking AI information seriously.” OpenAI was formed arsenic a counterweight to that threat, Musk claimed.