In my long list of Story Ideas I’ve Never Gotten Around To Writing is a Cold War (2) Spy Thriller centered on rival AI spies, one from the US/Europe, which has degenerated into a Snowcrash style dystopia where the AIs control everything by keeping people isolated and the other from China, which has degenerated to a 1984 style surveillance state in order to Keep AI from taking over.
Maybe I should write it before it’s no longer speculative…
Is the chaos brought about by a dangerous AI likely to be confined to one country? Or would it easily skip over national borders? If the latter, then perhaps the chaos-in-the-US-but-not-in-China scenario is less likely than made out to be. I do not know the answer, though. To some extent, the very notion of an AI arms race supposes that the impact of AI can be kept (fairly) local.
Depends on the nature of the chaos. Disruption of the job market, for example, could differ significantly among nations depending on how they regulate uses of AI.
Honestly, I'd say the concept of "Cold War with China" is nothing more than manufactured warhawk/military industrial complex propaganda. This isn't to say China has no threats (rampant IP theft, imperialist expansion, etc.), but these are petty compared to how inflammatory and combative the US is being.
Actually, maybe Cold War is the perfect description, as it was also manufactured and exacerbated by US aggression.
I would encourage Bob to check out the video game, Horizon: Zero Dawn to see an realistic situation of a bad actor creating A.I. that unintentionally destroy humanity.
Basically it's where the militaries of all nations turned to a.i. powered armies of machines. All controlled by a rich guy that accidentally patched the machine control software so it couldn't be accessed (no more human control) and then they destroyed everything.
From the wiki - the Old World was destroyed nearly a thousand years ago after Faro (rich guy company) lost control of its automated peacekeeper military robots. The robots, which could self-replicate and consume biomass, overran the planet and engulfed the biosphere, stripping Earth of life.
I would also say anyone should check out “Friendship Is Optimal” for a plausible scenario of an AI researcher making the best of a world where the AI takeover is inevitable.
(I've been in the AI/ neuro game for about 50 years (at MIT, UCSF, and Stanford).
1) I agree with your comments about Xi Jinping and the CCP; they are total control freaks! It's all about surveillance (the main thrust of the PRC's AI efforts) and keeping the citizenry in line.
2) In re: your notion that America will spin out of control with LLMs: what's currently out of control is Russian and Chinese troll farms (look at comments on YouTube); Meta's review board is great but those poor humans can only handle maybe a dozen issues per year. How to handle the rest of the circulating misinfo? The answer is "white hat" AGIs acting as smart spam filters.
3) In re: America as a valued market for Chinese goods; yeah, partially true, but that doesn't explain their massive military buildup (hugely expanded arsenal of ICBMs, subs, destroyers, etc)
The author is dreaming: no government or combination of governments can stop or even significantly slow the progress of AI.
"America is now turning itself into the canary in the AI coal mine"
Employing purple prose does nothing to put any substance into the fear of AI infecting the author and so many others. Just because a computer program is "smarter" than actual humans doesn't mean it runs our lives: we can still pull the plug or just ignore it.
I liked this one. Brings up some good view points to think about. To my view point the world needs to look at things we can work on together to improve world conditions everywhere. These are areas that we can come together on and create peace at the same time. Enterally in the US we have states that have almost third world conditions. No wonder we are so divided. The capital of one state is still boiling water. And it goes on. It would be hard for other countries to take us seriously when we let these unequal conditions go on in our country.
It’s kinda like it’s business opponents are in sync…’we get to collect your data, but not those guys’…because communism??
In my long list of Story Ideas I’ve Never Gotten Around To Writing is a Cold War (2) Spy Thriller centered on rival AI spies, one from the US/Europe, which has degenerated into a Snowcrash style dystopia where the AIs control everything by keeping people isolated and the other from China, which has degenerated to a 1984 style surveillance state in order to Keep AI from taking over.
Maybe I should write it before it’s no longer speculative…
Is the chaos brought about by a dangerous AI likely to be confined to one country? Or would it easily skip over national borders? If the latter, then perhaps the chaos-in-the-US-but-not-in-China scenario is less likely than made out to be. I do not know the answer, though. To some extent, the very notion of an AI arms race supposes that the impact of AI can be kept (fairly) local.
Depends on the nature of the chaos. Disruption of the job market, for example, could differ significantly among nations depending on how they regulate uses of AI.
Honestly, I'd say the concept of "Cold War with China" is nothing more than manufactured warhawk/military industrial complex propaganda. This isn't to say China has no threats (rampant IP theft, imperialist expansion, etc.), but these are petty compared to how inflammatory and combative the US is being.
Actually, maybe Cold War is the perfect description, as it was also manufactured and exacerbated by US aggression.
I would encourage Bob to check out the video game, Horizon: Zero Dawn to see an realistic situation of a bad actor creating A.I. that unintentionally destroy humanity.
I have to actually play the game to get the picture? That's a big ask...
I think you can get the idea just from the wiki or https://www.pushsquare.com/guides/horizon-zero-dawn-full-story-recap#before-horizon-zero-dawn
Basically it's where the militaries of all nations turned to a.i. powered armies of machines. All controlled by a rich guy that accidentally patched the machine control software so it couldn't be accessed (no more human control) and then they destroyed everything.
From the wiki - the Old World was destroyed nearly a thousand years ago after Faro (rich guy company) lost control of its automated peacekeeper military robots. The robots, which could self-replicate and consume biomass, overran the planet and engulfed the biosphere, stripping Earth of life.
As my mother used to say, never let a rich guy control your killer bots.
I would also say anyone should check out “Friendship Is Optimal” for a plausible scenario of an AI researcher making the best of a world where the AI takeover is inevitable.
https://www.fimfiction.net/story/62074/2/friendship-is-optimal/1-opportunity
(yes, it’s a My Little Pony fanfic, but you don’t have to be a brony to appreciate it)
I disagree with some of your points. My detailed argument is posted here:
https://www.bobblum.com/ESSAYS/COMPSCI/pause-llm.html
(I've been in the AI/ neuro game for about 50 years (at MIT, UCSF, and Stanford).
1) I agree with your comments about Xi Jinping and the CCP; they are total control freaks! It's all about surveillance (the main thrust of the PRC's AI efforts) and keeping the citizenry in line.
2) In re: your notion that America will spin out of control with LLMs: what's currently out of control is Russian and Chinese troll farms (look at comments on YouTube); Meta's review board is great but those poor humans can only handle maybe a dozen issues per year. How to handle the rest of the circulating misinfo? The answer is "white hat" AGIs acting as smart spam filters.
3) In re: America as a valued market for Chinese goods; yeah, partially true, but that doesn't explain their massive military buildup (hugely expanded arsenal of ICBMs, subs, destroyers, etc)
Bob Blum, MD, PhD (Stanford)
Thanks , cool post! Refreshingly clear and not biased by anti-China ism
The author is dreaming: no government or combination of governments can stop or even significantly slow the progress of AI.
"America is now turning itself into the canary in the AI coal mine"
Employing purple prose does nothing to put any substance into the fear of AI infecting the author and so many others. Just because a computer program is "smarter" than actual humans doesn't mean it runs our lives: we can still pull the plug or just ignore it.
I liked this one. Brings up some good view points to think about. To my view point the world needs to look at things we can work on together to improve world conditions everywhere. These are areas that we can come together on and create peace at the same time. Enterally in the US we have states that have almost third world conditions. No wonder we are so divided. The capital of one state is still boiling water. And it goes on. It would be hard for other countries to take us seriously when we let these unequal conditions go on in our country.