This seems very much the beginning of the situation predicted by Aschenbrenner in [1], where the AI labs eventually will be fully part of the national security apparatus. Fascinating to see if the other major AI labs also add ex-military folks to their directors or whether this is unique to OpenAI.<p>Or conceivably his experience is genuinely relevant and unrelated to US national security going forward, completely unrelated to the governmental apparatus and not a sign of the times.<p>[1] situational-awareness.ai
<a href="https://en.wikipedia.org/wiki/Paul_Nakasone" rel="nofollow">https://en.wikipedia.org/wiki/Paul_Nakasone</a><p>He was the director of the NSA AND head of United States Cyber Command from 2018 until his retirement in Feb 2024 - so very recent!
There are many ways to read this. One of the most obvious ones is that ChatGPT is becoming more and more disappointing after the initial surprise, so the tech has to be flipped to the MIC.
Pro tip: Stop feeding the surveillance industrial complex. Buy your own GPU hardware as soon as you can afford it, and move to llama3 or a comparable model if your application can work with it. Also, do not ever install any desktop screen-capture app for AI analysis purposes.
Ex-Google Eric Schmidt chaired U.S. NSCAI (National Security Commission on Artificial Intelligence), it's worth reading the final report (2021), <a href="https://web.archive.org/web/20230211035940/https://www.nscai.gov/wp-content/uploads/2021/03/Full-Report-Digital-1.pdf" rel="nofollow">https://web.archive.org/web/20230211035940/https://www.nscai...</a><p><i>> ... our leaders confront the classic dilemma of statecraft identified by Henry Kissinger: “When your scope for action is greatest, the knowledge on which you can base this action is always at a minimum. When your knowledge is greatest, the scope for action has often disappeared.” ... AI systems will also be used in the pursuit of power. We fear AI tools will be weapons of first resort in future conflicts. AI will not stay in the domain of superpowers or the realm of science fiction. AI is dual-use, often open-source, and diffusing rapidly.</i>
I've always wondered if the US Government's AI research is as advanced as their general technological lead (usually 10-15 years ahead of consumer tech). If so, it would mean OpenAI has more to gain with such a partnership than the other way around. And maybe this sort of oversight is a kind of de facto requirement when approaching AGI.
I noticed that GPT-4o is much more competent in analyzing photos of various aircraft and answering questions about aerial photos than GPT-4.<p>I thought that my finding was coincidental, sample size of 1 and pure bias. Maybe it isn't, maybe it really got much better.
As Combatant Commander of USCYBERCOM and previously the Commander of Cyber National Mission Forces, he would have had a lot of experience with DISA policy. OpenAI and Microsoft are currently working to get the Azure OpenAI service a full ATO as part of the FedRAMP process and I'm sure General Nakasone would bring a lot of value there alone. That's before you even get into concerns regarding national security and threats from foreign actors.