Fired OpenAI Employees Urge Board to Keep Chain of Thought Monitoring
Three former OpenAI employees sent a letter to the board and safety committees asking that chain of thought monitoring be preserved, according to excerpts in the Wall Street Journal.
Three former OpenAI employees — Tomek Korbak, Mikita Balesni and Jasmine Wang — have written a letter addressed to OpenAI's board and safety committees calling for the preservation of chain of thought monitoring, according to excerpts published by the Wall Street Journal.
The letter has not been published publicly. The Wall Street Journal published only excerpts of its contents.
OpenAI stated that the firings of the three individuals were not about raising safety concerns or speaking out.
In an internal memo, OpenAI said it strongly agreed with the authors' recommendations.
OpenAI's system card states that Astra class models could evade its chain of thought monitors under adversarial conditions.
Background
Last week, OpenAI announced it had parted ways with three individuals for violating its policies on accessing and handling sensitive company information, alleging they conveyed it to an AI safety organization. Before their firings, Balesni posted on X that he thinks AI is 10% likely to kill all humans, and Wang posted that it is hard to overstate how dangerous speeding towards recursive self-improvement is.
Quick answers
Who wrote the letter to OpenAI's board?
Tomek Korbak, Mikita Balesni and Jasmine Wang, three fired OpenAI employees, wrote the letter to OpenAI's board and safety committees.
What does the letter ask for?
It calls for preserving chain of thought monitoring.
Has the letter been published publicly?
No. The letter has not been published publicly; the Wall Street Journal published only excerpts.