Or they received a classified briefing showing how far ahead China or other adversaries are in terms of AI for military use and felt obligated to not let the US fall further behind.

Why is this being downvoted? This is totally possible, though I'm not saying it happened.

Is it because it's speculative?

Hard to say why it was downvoted but I can say it’s a highly unlikely scenario. You typically need a security clearance to view classified info.

Further, many employees at OpenAI may not even be US citizens, and even if they are, many others aren’t super enthusiastic about giving the Pentagon new toys to use.

OpenAI had/has 57 Chinese talent [1], they're probably getting the boot if OpenAI picks up Pentagon contracts.

https://macropolo2.wpenginepowered.com/wp-content/uploads/20...

For dual-use tech, its relatively common to have a small "government" division alongside the much larger commercial division. If the only difference for government is fine tuning data/DoD servers - it shouldn't cost OpenAI much to support.

People are escalating AI to strategic/end of world tier of dual use tech. Questionable if it will be treated the same way. OpenAI is not lockheed, but question is how much they value Pentagon contracts or really gov connections especially with all the regulation talk.

There will certainly be DARPA and other grants to test if dedicated defense suppliers make sense. In some areas, the funding model may be easier as a pure defense initiative where the DoD could shell out 50-100 Billion for cross-service AI functionality.

In theory there is no reason why an individual soldier shouldn't be able to order an air strike, or be informed about inbound strikes on their position and how to react to those. Today, this is not possible due to the complex decisioning process for whether a target is worth a strike by X munition vs. Y munition and when. AI could simplify this process.

On the other hand, well funded open source and commercial companies may deliver faster and with higher quality. The DoD doesn't use bespoke databases, and microchips for a reason (except when they do).

how did you even acquire this graph?

That would be a great loss of talent. I wish we could make it much easier for those top scientists to become American citizens.

Why assume they have loyalty to the CCP? They’re in California, presumably they like hot dogs, monster trucks, and bikinis.

They also like their friends and family members alive and well.

> Why assume they have loyalty to the CCP?

You answered

> They’re in California

>Why assume they have loyalty to the CCP

Because that's American NatSec for you. They rather train a white kid from Iowa to read Chinese poorly and do even worse analysis than risk giving advantage to adversaries. The domestic politics of latter is more unforgiving. Also see China Initiative persecuting PRC scholars. Pentagon contracts will come with strings attached, hidden or otherwise.

I can't count the number of times I have seen Chinese people abroad having their family back at home punished for criticizing the CCP or taking other stances that the government oppose.

Doing a quick search, I found several articles detailing how this is becoming more common, with this story being the first one I recalled as the video is nauseating: https://www.voanews.com/a/east-asia-pacific_china-tries-muff... "Editor’s Note: China’s government is expanding its censorship controls by targeting Chinese citizens overseas who criticize Beijing on social media. The pressure tactic is called “zhulian” – an ancient punishment meaning “guilt by association.” It usually involves police threatening family members in China for the actions of their relatives overseas.

This happened to “Zoo” (short for @HorrorZoo, her main handle on Twitter), a Chinese student pursuing graduate studies in Australia, whose father has been repeatedly summoned to the police station because of her criticism of Chinese Communist Party leaders on social media.

But instead of censoring her Twitter account, “Zoo” has become more outspoken, providing multiple videos of police intimidating her via video chat. Her story provides a rare look at how China tries to use the families of dissidents to silence them."

Another instance: https://www.rfa.org/english/news/china/overseas-chinese-1021... "Police in Beijing have contacted the family of a Chinese student studying in the United States after he expressed support online for the "Bridge Man" protester, who unfurled banners on a Beijing bridge calling on ruling Chinese leader Xi Jinping to step down on the eve of the 20th Communist party’s congress."

I'm sure I could find hundreds of these - it has nothing to do with loyalty, and everything to do with external sources of pressure and coercion.

Hey yeah maybe we could put them all into one place to keep an eye on them easier

Like taking people that are all over the place and sort of concentrating them into one place

Like a camp

Like a concentration camp

For Chinese people

You know because we can’t trust them not to be traitors

That seems like a good idea right

WMDs anyone?

Weapons of Math Destruction?

A good book I recommend to read

+1

I read this in Mike Tyson's voice.

Perhaps they were put on a bus, taken to Graceland and met the actual Elvis who is still alive and well sitting around jamming with JFK.

Last I heard the Chinese had found a way of fueling their rockets with tap water. We're screwed regardless of their AI prowess.

I just spit sparkling water through my nose and it’s like 3am and IM in bed.

I am sure many employees signed the letter out of loyalty. Not sure Altman is being very loyal to them in return.

Persuasion is often a strong skill in managers. Personal motivation absent outside forces has its work cut out for it in corporations.

Observing Altman over OpenAI saga and later I clearly see one thing: duplicity. He was kicked off YC by Graham for unknown reasons, yet after October fiasco I have my suspicions why. Coming back to OpenAI and its beginnings, Altman betrayed all early investors - notably Musk but not only. Betrayed each and every meaningful point of OpenAI original charter, basically making it the oposite of what it originally was - starting exactly at the point when he smelled big money. He even went to Congress, effectively lobbying for limiting competition - especially open source models. Now he betrayed humanitarian values by offering his services for Pentagon - organization that has killed untold millions of people worldwide since WW2 and will happily kill millions more, making a gobs money to MIC in the process (don't tell me about "democracy" and such, it's all bullshit). There is no reason to believe a single word from him, Altman always tells his agenda, nothing else. There will be no 'using AGI for helping humanity' - his corporate buddies will receive profits at the cost of everyone else losing job to AI solutions monopolized by Altman's buddies. There will be no 'distributing profits among people in need', maybe except for some token help just to shut up critics. Altman is a tool for ruling corporate oligarchy. Nothing good will come out of it in my opinion.

We are so back! :heart emoji: :heart emoji: :heart emoji:

Deeply deeply concerning.

You’re assuming these are separate events.

Close. I think reinstating Altman was all about this and also maybe the money.

What better explanation do we have for the violent reaction of the board? I read the cover story and analysis too, but there are no coincidences in National Security.

Occam’s Razor points to this being the real reason.