the lizard man spent 6,500 words to essentially say "trust us to distribute power responsibly" And yet it never mentions the last time this company distributed something at scale. No mention of Cambridge Analytica, no Myanmar, no engagement optimization, no youth harm litigation. And I never saw the word "advertising" appear. Not to mention that he argues against centralized AI while describing an undertaking that maybe five entities on earth can finance. The things we need to make this possible, the chips, the energy, the capital, are already centralized. It doesn't matter how many people you hand the technology to. I could give you 6,500 words of just the things Zuckerberg left out.
First, what Zuck says cannot be taken at face value, of course. The only difference between him and Altman or Google's leadership is that he is behind in the tech. He is precisely in the same mold, going back to Facebook's founding. Say whatever it takes to get ahead, and switch track as soon as is convenient.
Aso to AI for everyone, the idea is nice, in principle. In practice I see no good way of getting this off the ground. We will have some control over our own AI, as we do nowadays over our smartphone software and our privacy. It will go however as far as corporations and (most important) governments will allow.
Truly independent, home-grown, decentralized, all-yours solution is not more practical than your own off-the-grid electricity, water supply, or a server in basement. Possible in some special cases, not in general. And governments won't allow it for security reasons even if feasible.
I will add to your comment that not everyone wants AGI/ASI in the political/economic/social sphere at all. It is like arguing that nuclear weapons are for everyone, but some countries have eschewed their acquisition. If Russia used a nuke on Ukraine or on any NATO country, you can bet all NATO countries would be forced to have their own too. The world would be better off not having any nukes with the temptation to use them. The same may apply to very powerful, potentially uncontrollable AI.
Iain Banks' Culture universe sci-fi novels may seem benign for the populations, but I would find it as stultifying as having a single religion with its deity in control. We have seen where that led in Europe for over a millennium. No thank you. {And we can see what the Christo-Fascists trying to take the USA in that direction will lead to.]
AI sadly is not something you will be able to turn off globally. You can go off-grid, as you do with other things, such as living in a tent, or without internet, etc., but for society as a whole it is a given.
The Culture is likely the best we can hope for. The AI there gave everybody maximal freedom, up to the point of choosing death, crazy sports, bizarre hobbies, and ability to escape it altogether and joining other civilizations and ancient alien ecosystems. Most people were happy to do their own little thing within the AI matrix, blissfully clueless about how it was all managed.
> The core of the essay is an argument against what we might call the singleton model of AI safety — the notion that the path to a beneficial superintelligence is for one lab to build one system, align it very carefully with one set of values, and then let it benevolently manage things on humanity’s behalf. Zuck argues, correctly, that this is not a reliable path to benevolence — that humanity is not a value monoculture, that any singular superintelligence would have to privilege some people’s values over others, and that concentrations of absolute power have historically not worked out well no matter how enlightened the folks at the top believed themselves to be. What we need instead is diversity — diversity of methods, diversity of value systems, a complex dynamic cognitive and value ecology.
In 2014 I tried to recruit E.O. Wilson on a grant proposal to study how to apply principles from sociobiology and possibly group selection to design an evolutionarily stable ecology of altruistic AIs. He was too busy to take on another grant, but he called me up to say he liked the idea, and we talked for half an hour. About what, I have no idea. I was too much in awe of him to remember to remember.
One only has to look at how Zuck runs Facebook. It's a train wreck with a terrible reputation for respecting privacy, or providing customer support when deplatformed for unspecified or incorrect reasons. Why would Zuck do any better when providing AIs to customers? I wouldn't touch it with a 10 ft pole. He's the LAST person I'd want to wield that kind of power over people.
The New Yorker article is slanted and one sided. It says "Trump started a trade war (with China)." Total nonsense. China was trading very unfairly with the US. Trump pointed that out and is trying to even the playing field so the United States is no longer being taken advantage of. That's not "starting a trade war." It's responding appropriately in the best interests of the United States. The article makes no mention of the fact that AI technology, including LLMs and semiconductors, was created in the US, NOT China. China steals intellectual property; it rarely creates it. The US has an entrepreneurial culture, unlike China, that gives it an enormous advantage. Sure, China has made enormous strides and is catching-up, but the article makes it sound as if China is ahead of the US technologically, or soon will be. The article also makes it sound as though China's tech superiority is inevitable, which is ridiculous.
Dang, Dawg, you are The Shit. I got onto you via David Eagleman. I will be subscribing. I’ve spent all day reading “The Infinity Machine“ about Habbis & DeepMind.
the lizard man spent 6,500 words to essentially say "trust us to distribute power responsibly" And yet it never mentions the last time this company distributed something at scale. No mention of Cambridge Analytica, no Myanmar, no engagement optimization, no youth harm litigation. And I never saw the word "advertising" appear. Not to mention that he argues against centralized AI while describing an undertaking that maybe five entities on earth can finance. The things we need to make this possible, the chips, the energy, the capital, are already centralized. It doesn't matter how many people you hand the technology to. I could give you 6,500 words of just the things Zuckerberg left out.
First, what Zuck says cannot be taken at face value, of course. The only difference between him and Altman or Google's leadership is that he is behind in the tech. He is precisely in the same mold, going back to Facebook's founding. Say whatever it takes to get ahead, and switch track as soon as is convenient.
Aso to AI for everyone, the idea is nice, in principle. In practice I see no good way of getting this off the ground. We will have some control over our own AI, as we do nowadays over our smartphone software and our privacy. It will go however as far as corporations and (most important) governments will allow.
Truly independent, home-grown, decentralized, all-yours solution is not more practical than your own off-the-grid electricity, water supply, or a server in basement. Possible in some special cases, not in general. And governments won't allow it for security reasons even if feasible.
I will add to your comment that not everyone wants AGI/ASI in the political/economic/social sphere at all. It is like arguing that nuclear weapons are for everyone, but some countries have eschewed their acquisition. If Russia used a nuke on Ukraine or on any NATO country, you can bet all NATO countries would be forced to have their own too. The world would be better off not having any nukes with the temptation to use them. The same may apply to very powerful, potentially uncontrollable AI.
Iain Banks' Culture universe sci-fi novels may seem benign for the populations, but I would find it as stultifying as having a single religion with its deity in control. We have seen where that led in Europe for over a millennium. No thank you. {And we can see what the Christo-Fascists trying to take the USA in that direction will lead to.]
AI sadly is not something you will be able to turn off globally. You can go off-grid, as you do with other things, such as living in a tent, or without internet, etc., but for society as a whole it is a given.
The Culture is likely the best we can hope for. The AI there gave everybody maximal freedom, up to the point of choosing death, crazy sports, bizarre hobbies, and ability to escape it altogether and joining other civilizations and ancient alien ecosystems. Most people were happy to do their own little thing within the AI matrix, blissfully clueless about how it was all managed.
> The core of the essay is an argument against what we might call the singleton model of AI safety — the notion that the path to a beneficial superintelligence is for one lab to build one system, align it very carefully with one set of values, and then let it benevolently manage things on humanity’s behalf. Zuck argues, correctly, that this is not a reliable path to benevolence — that humanity is not a value monoculture, that any singular superintelligence would have to privilege some people’s values over others, and that concentrations of absolute power have historically not worked out well no matter how enlightened the folks at the top believed themselves to be. What we need instead is diversity — diversity of methods, diversity of value systems, a complex dynamic cognitive and value ecology.
In 2014 I tried to recruit E.O. Wilson on a grant proposal to study how to apply principles from sociobiology and possibly group selection to design an evolutionarily stable ecology of altruistic AIs. He was too busy to take on another grant, but he called me up to say he liked the idea, and we talked for half an hour. About what, I have no idea. I was too much in awe of him to remember to remember.
Goertzel.org dns has expired
See SingularityNET.io
One only has to look at how Zuck runs Facebook. It's a train wreck with a terrible reputation for respecting privacy, or providing customer support when deplatformed for unspecified or incorrect reasons. Why would Zuck do any better when providing AIs to customers? I wouldn't touch it with a 10 ft pole. He's the LAST person I'd want to wield that kind of power over people.
BTW, you might dig this: https://www.newyorker.com/magazine/2026/08/10/the-future-made-in-china
The New Yorker article is slanted and one sided. It says "Trump started a trade war (with China)." Total nonsense. China was trading very unfairly with the US. Trump pointed that out and is trying to even the playing field so the United States is no longer being taken advantage of. That's not "starting a trade war." It's responding appropriately in the best interests of the United States. The article makes no mention of the fact that AI technology, including LLMs and semiconductors, was created in the US, NOT China. China steals intellectual property; it rarely creates it. The US has an entrepreneurial culture, unlike China, that gives it an enormous advantage. Sure, China has made enormous strides and is catching-up, but the article makes it sound as if China is ahead of the US technologically, or soon will be. The article also makes it sound as though China's tech superiority is inevitable, which is ridiculous.
https://youtu.be/7LnNrcLc5K0?is=ptwQmnBj_148j8pm
“The article makes no mention of the fact that AI technology, including LLMs and semiconductors, was created in the US, NOT China.”
Yeah. Exclusively by young Silicon Valley white male bro’ U.S. citizens with names like Smith, Jones, Williams, Rogers, Stevens, Allan, ……..
Andy Grove, Gordon Moore, Steve Wozniak... you mean like THOSE last names? All AMERICAN, as well as the one's who's families did not come from Europe.
Dang, Dawg, you are The Shit. I got onto you via David Eagleman. I will be subscribing. I’ve spent all day reading “The Infinity Machine“ about Habbis & DeepMind.