Well done Bob. I thoroughly enjoyed your book and wrote about it and subsequently about AI several other times, largely inspired by your book and others. It seems that AI (not energy, not nuclear weapons, not global climate change, not global pandemics and/or bioweapons) might, might just finally be the culminating technology that brings the world around to Bob W. and non-zero sum thinking, and hey maybe even progressive realism to boot, all in service to and the recognition that we live on one blue planet and "we" better get it together or "we" will not like the result. Keep on pushing on this, your time has arrived.
i'm with you that the singularity is clear, and i'd push it one turn further. the part that's clear is the mechanism: acceleration, feedback, the menu of governance outcomes. what stays genuinely opaque isn't the future's shape, it's whether we remain the kind of beings who can choose it.
good global governance assumes a public that can still attend, deliberate, want the good. but the same technology bending the geopolitics is also bending the interior: attention, judgment, the capacity for a self that isn't just optimized.
the real event horizon might not be political. it might be anthropological. we can see the worlds available to us. the open question is whether we'll still be the sort of creatures capable of picking one.
Any society that elected Trump cannot attend, deliberate and want the good. Social media has become an uncontrolled destroyer of intellect and yet we let it grow without controls. Why would AI be any different ? Are there truly any Leaders that can lead us in the right direction ? Certainly not any Politicians. Tech Leaders, doubtful. Yes, The event horizon will be anthropological / spiritual and perhaps be led by AI itself.
I think the given choices for structuring our future are correct, but your adjective does a ton of work. Surely "Good" global governance has to mean something more specific than competent coordination among whoever holds the chips; I can think of several regimes that ran efficiently but that we would call "Bad".
I'm convinced that "Good" global governance can only mean an order where worth isn't differentially protected, especially on arbitrary criteria like race, or gender, or what lines on the Mercator projection you were born inside. Egalitarian moral worth implies that our institutions should treat humans as such, and protect lives equally.
A singleton run by a concert of states that still guarantees behind-birthplace guarantees is not that, and I wouldn't call it "Good" no matter how efficient. Why not specify the political choice you describe? I'd go from simply (governance vs chaos) to (a global polity with a universal floor vs a global cartel of nations). The first one seems falsifiably good.
"So here is the full menu of options: (1) good global governance; (2) bad global governance; (3) an enduringly chaotic hellscape".
Option 1 is out of the question - it assumes for example that all the jihadis, communists, fascists, psychopaths etc have somehow disappeared.
Option 2, assuming global governance is in place, human nature could be relied upon that everyone would regard it bad in some respect - so unfalsifiable.
Option 3, if it was a hellscape then it wouldn't endure - at least not as a global governance.
There is an option 4 which is arguably more likely. A global governance with limited powers that muddles along, good enough not to piss everyone off so much that it triggers a revolution whilst nevertheless still causing some terrible blunders from time to time.
Beyond that, the case for AI ruling the world seems way over blown. As a tool for amplifying the abilities of skilled and talented people AI unquestionably has a bright future. As has always been the case in the history of automation, there will be many entry and middle level jobs which will be replaced by AI. What is always harder to predict are the opportunities for new and better jobs that will inevitably open up as a result.
What should not be forgotten is that AI is not really intelligence at all it is really just a sophisticated predictive text mechanism - it has no real understanding of anything nor is there any evidence of it being anywhere near having a real understanding. This contrasts with human cognition which has arisen out of evolutionary selection over millions of years. To assume that a computer algorithm is going to out perform that at the most difficult levels of cognition is hubris to say the least.
That's not to say AI couldn't cause some terrible disasters. It's just that the idea that it would consciously do so in anything other than a metaphorical sense is good as impossible in the foreseeable future.
A good international order does not necessarily require a governing body , a harmonious interaction of rational nation states can achieve a reasonable result and a coalition of good guys can reign in bad ones
I read the book. Very good distillation of the landscape. I highly recommend it.
Picky point, Ulam didn't co-invent game theory. That was Morgenstern.
Von Neumann said that Edward Teller was the father of the H bomb, but Ulam slept with its mother.
good point. thanks. i'll make that correction.
Well done Bob. I thoroughly enjoyed your book and wrote about it and subsequently about AI several other times, largely inspired by your book and others. It seems that AI (not energy, not nuclear weapons, not global climate change, not global pandemics and/or bioweapons) might, might just finally be the culminating technology that brings the world around to Bob W. and non-zero sum thinking, and hey maybe even progressive realism to boot, all in service to and the recognition that we live on one blue planet and "we" better get it together or "we" will not like the result. Keep on pushing on this, your time has arrived.
https://tinyurl.com/AI-and-Kosmic-Reckoning
https://tinyurl.com/AI-Wisdom-and-Flourishing
i'm with you that the singularity is clear, and i'd push it one turn further. the part that's clear is the mechanism: acceleration, feedback, the menu of governance outcomes. what stays genuinely opaque isn't the future's shape, it's whether we remain the kind of beings who can choose it.
good global governance assumes a public that can still attend, deliberate, want the good. but the same technology bending the geopolitics is also bending the interior: attention, judgment, the capacity for a self that isn't just optimized.
the real event horizon might not be political. it might be anthropological. we can see the worlds available to us. the open question is whether we'll still be the sort of creatures capable of picking one.
Any society that elected Trump cannot attend, deliberate and want the good. Social media has become an uncontrolled destroyer of intellect and yet we let it grow without controls. Why would AI be any different ? Are there truly any Leaders that can lead us in the right direction ? Certainly not any Politicians. Tech Leaders, doubtful. Yes, The event horizon will be anthropological / spiritual and perhaps be led by AI itself.
I think the given choices for structuring our future are correct, but your adjective does a ton of work. Surely "Good" global governance has to mean something more specific than competent coordination among whoever holds the chips; I can think of several regimes that ran efficiently but that we would call "Bad".
I'm convinced that "Good" global governance can only mean an order where worth isn't differentially protected, especially on arbitrary criteria like race, or gender, or what lines on the Mercator projection you were born inside. Egalitarian moral worth implies that our institutions should treat humans as such, and protect lives equally.
A singleton run by a concert of states that still guarantees behind-birthplace guarantees is not that, and I wouldn't call it "Good" no matter how efficient. Why not specify the political choice you describe? I'd go from simply (governance vs chaos) to (a global polity with a universal floor vs a global cartel of nations). The first one seems falsifiably good.
"So here is the full menu of options: (1) good global governance; (2) bad global governance; (3) an enduringly chaotic hellscape".
Option 1 is out of the question - it assumes for example that all the jihadis, communists, fascists, psychopaths etc have somehow disappeared.
Option 2, assuming global governance is in place, human nature could be relied upon that everyone would regard it bad in some respect - so unfalsifiable.
Option 3, if it was a hellscape then it wouldn't endure - at least not as a global governance.
There is an option 4 which is arguably more likely. A global governance with limited powers that muddles along, good enough not to piss everyone off so much that it triggers a revolution whilst nevertheless still causing some terrible blunders from time to time.
Beyond that, the case for AI ruling the world seems way over blown. As a tool for amplifying the abilities of skilled and talented people AI unquestionably has a bright future. As has always been the case in the history of automation, there will be many entry and middle level jobs which will be replaced by AI. What is always harder to predict are the opportunities for new and better jobs that will inevitably open up as a result.
What should not be forgotten is that AI is not really intelligence at all it is really just a sophisticated predictive text mechanism - it has no real understanding of anything nor is there any evidence of it being anywhere near having a real understanding. This contrasts with human cognition which has arisen out of evolutionary selection over millions of years. To assume that a computer algorithm is going to out perform that at the most difficult levels of cognition is hubris to say the least.
That's not to say AI couldn't cause some terrible disasters. It's just that the idea that it would consciously do so in anything other than a metaphorical sense is good as impossible in the foreseeable future.
"As you approach its center (which isn’t recommended)" - true!
or where, like one of those vortexes from Verne's sci-fi stories, our curiosity would dictate us to 'boldly go where no-one's gone before' :-)
Someone has to!
A good international order does not necessarily require a governing body , a harmonious interaction of rational nation states can achieve a reasonable result and a coalition of good guys can reign in bad ones