People with previous and current roles at OpenAI and Google DeepMind referred to as for the safety of critics and whistleblowers on June 4.
Authors of an open letter urged AI firms to not enter agreements that block criticism or retaliate in opposition to criticism by hindering financial advantages.
Moreover, they acknowledged that firms ought to create a tradition of “open criticism” whereas defending commerce secrets and techniques and mental property.
The authors requested firms to create protections for present and former staff the place current danger reporting processes have failed. They wrote:
“Bizarre whistleblower protections are inadequate as a result of they deal with criminal activity, whereas most of the dangers we’re involved about are usually not but regulated.”
Lastly, the authors mentioned that AI corporations ought to create procedures for workers to lift risk-related considerations anonymously. Such procedures ought to permit people to lift their considerations to firm boards and exterior regulators and organizations alike.
Private considerations
The letter’s 13 authors described themselves as present and former staff at “frontier AI firms.” The group consists of 11 previous and current members of OpenAI, plus one previous Google DeepMind member and one current DeepMind member, previously at Anthropic.
They described private considerations, stating:
“A few of us moderately worry numerous types of retaliation, given the historical past of such circumstances throughout the business.”
The authors highlighted numerous AI dangers, equivalent to inequality, manipulation, misinformation, lack of management of autonomous AI, and potential human extinction.
They mentioned that AI firms, together with governments and consultants, have acknowledged dangers. Sadly, firms have “sturdy monetary incentives” to keep away from oversight and little obligation to share personal details about their programs’ capabilities voluntarily.
The authors in any other case asserted their perception in the advantages of AI.
Earlier 2023 letter
The request follows an April 2023 open letter titled “Pause Big AI Experiments,” which equally highlighted dangers round AI. The sooner letter gained signatures from business leaders equivalent to Tesla CEO and X chairman Elon Musk and Apple co-founder Steve Wozniak.
The 2023 letter urged firms to pause AI experiments for six months in order that policymakers might create authorized, security, and different frameworks.