A former OpenAI employee is urging prominent law enforcement officials in California and Delaware to halt the transition of the company’s artificial intelligence management from non-profit organizations to commercial enterprises.
Concerns have been raised regarding the implications if the makers of ChatGPT fulfill their ambition to develop AI that surpasses human capabilities while evading accountability for their public mission of preventing potential disastrous consequences from their technology.
“Ultimately, I’m concerned about who will possess and manage this technology once it’s created,” stated Page Hedley, ex-policy and ethics advisor for OpenAI, in an interview with the Associated Press.
Backed by three Nobel Prize winners, along with various supporters and experts, Hedley and nine other former OpenAI employees have sent a letter to two state attorneys general this week.
This coalition is calling on both Democrats, California Attorney General Rob Bonta and Delaware Attorney General Kathy Jennings, to leverage their powers to safeguard open charitable purposes and prevent the intended restructuring. OpenAI is incorporated in the state of Delaware and operates out of San Francisco.
“Modifications to existing frameworks will allow the broader public to benefit from AI,” OpenAI stated. Although they maintain that they will evolve into a public benefits corporation, like other AI laboratories such as Humanity and tech billionaire Elon Musk’s X.ai, OpenAI continues to uphold its non-profit sector.
“This structure is designed to guarantee that as for-profit entities prosper and expand, non-profits will also grow and fulfill their missions,” the company emphasized in a statement.
The letter represents the second petition lodged this month against state officials. The first was initiated by a group of labor leaders and non-profits highlighting OpenAI’s substantial protection of charitable assets.
Jennings mentioned last fall that he would “assess such transactions to ensure that public interests are duly safeguarded.” Bonta’s office sought additional information from OpenAI late last year but refrained from commenting on whether an investigation was underway or denying it.
OpenAI’s co-founders, including current CEOs Sam Altman and Musk, initially established it as a non-profit research institute with the mission to safely develop what is known as artificial general intelligence or AGI for the benefit of humanity. Nearly ten years later, OpenAI claims a market valuation of $300 billion, with 400 million weekly users of its flagship product, ChatGPT.
OpenAI already has for-profit entities; however, it faces numerous challenges in transforming its fundamental governance structure. One significant issue is a lawsuit from Musk, accusing the company and Altman of betraying the founding principles that prompted Tesla executives to invest in charitable causes.
While some signatories of this week’s letter support Musk’s legal actions, Hedley remarked that others are “fundamentally cynical” due to Musk operating his own rival AI company.
The letter’s signatories include Nobel Prize-winning economists Oliver Hart and Joseph Stiglitz, along with AI pioneer computer scientist Jeffrey Hinton, who was awarded the Nobel Prize in Physics last year, and Stuart Russell.
“I appreciate the open mission of ‘ensuring that artificial general intelligence benefits all of humanity’ and hope they will adhere to that mission rather than just enriching investors,” Hinton expressed in a statement on Wednesday. “We are pleased to observe efforts to maintain OpenAI’s mission that do not involve Elon Musk.”
The debate over OpenAI’s objectives has been ongoing at the San Francisco Institute, contributing to Musk’s resignation in 2018 and Altman’s brief removal in 2023, among other notable departures.
Hedley, who is a trained lawyer, was with OpenAI during 2017 and 2018, a time when non-profits were still exploring the best management approaches for the technology they aimed to develop. Recently, in 2023, Altman cautioned that advanced AI possesses immense potential yet also threatens significant risks, ranging from severe accidents to societal upheaval.
However, Hedley voiced concerns that OpenAI, buoyed by the success of ChatGPT, is compromising on safe testing and hastily launching new products to stay ahead of market competitors.
“The implications of these decisions are escalating as technology becomes more powerful,” he said. “I believe there is no one left to advise them against the new structure that OpenAI seeks, which incentivizes them to expedite these decisions.”
Software engineer Anish Tondwalkar, a former member of OpenAI’s technical team until last year, emphasized that a crucial safeguard in OpenAI’s non-profit charter was a “stop-assist clause”, which mandated OpenAI to pause and assist if another organization approached creating superior human-level AI.
“If OpenAI is allowed to transition into a for-profit organization, these safeguards, along with OpenAI’s obligations to the public, might vanish overnight,” Tondwalkar warned in a statement on Wednesday.
Another former employee who signed the letter candidly stated, “OpenAI may one day develop technology that could threaten our very existence,” said Nisan Stiennon, an AI engineer who worked with OpenAI from 2018 to 2020.
___
The Associated Press and OpenAI have a License and Technical Agreement that enables OpenAI to access certain archives from the AP’s text repository.
Source: apnews.com