The Entire OpenAI Explosion Was Ignited by This Research Paper
themessenger.com
themessenger.com
Read pages 1-3 for summary of the paper and pages 27-30 (under "private sector signaling") for comparison of OpenAI vs. Anthropic approaches to signaling.
Its use of the Cuban Missile crisis to compare to this moment AI is bizarre and alarmist. All for a concusion of... we must control the narrative. Spook juice.
First off, it is published by CSET whose mission is to look at "security implications of emerging technologies" and "to inform research at the intersection of technology and statecraft." (https://cset.georgetown.edu/about-us/) So the focus on Military/Geopolitics/Strategy is mandatory and natural.
The term "Costly Signal" is well known and used in Game Theory/International Relations/etc. fields (eg. https://www.youtube.com/watch?v=yfVTKG8O2bI&list=PLB5965C13F...) and thus has different aspects to its usage. The origin of the specific four used by the authors in the report is given by the very first citation in the "Endnotes" section.
The paper is titled "Decoding Intentions Artificial Intelligence and Costly Signals" and thus 1) the importance of figuring out the intention behind a signal that is being received and 2) the importance of clearly and accurately conveying intention on the sending side is what is being emphasized by the example of "Cuban Missile Crisis", and not that AI safety is of the same category as the "Cuban Missile Crisis".
In summary, everybody needs to read this report in entirety to really "grok" the importance and implications of "AI policy making".
Perhaps we are seeing the result of zero information, so like one inside of an isolation tank you have millions of people piecing together theories from what they can actually see, such as the fact that a paper exists by a board member and there's a sentence here or two which give a rationale.
IMO the dispute wasn't about anything in the paper unless Altman is crazy, which seems ruled out based on his fan base.
IMO it's about AGI. The board really doesn't want it and was willing to axe Altman because he is flirting with releasing a "dumbed down" version of it. This is under the assumption that AGI exists and is being kept secret, something most everyone thinks is totally insane.
Q* is powerless compared to Toner on that matter. Or, it didn't want to spend time on such a, admittedly fun, unrevealing and developing story. Q* knows?