That's why I say "accurate, extrapolatable-the-the-public evidence." With covid we ALSO coordinated shut down the world for varying durations on the basis of literally a chart showing a line going up to the right and some pictures from China.
So yes, I think my case that - if persuaded with sufficiently convincing evidence - we can coordi…
That's why I say "accurate, extrapolatable-the-the-public evidence." With covid we ALSO coordinated shut down the world for varying durations on the basis of literally a chart showing a line going up to the right and some pictures from China.
So yes, I think my case that - if persuaded with sufficiently convincing evidence - we can coordinate massive global action: works just fine. Obviously yes, things did not go perfectly and we did things we shouldn't have, and we did not do things we should have. The % of those errors we have for future events will depend on how accurately we understand the threat, and it's likelihood and proximity - we are in the "before the next disaster" phase right now. So get out there and find your "upwards to the right" chart, if it exists.
I’m not disputing your case for action — far from it. “Very little hope” is not the same as “despair”, and arguably should be more motivating than “somewhat more hope.” Unfortunately, most of my remaining hope on this would come from a scenario where some less-than-all-powerful AGI-esque thing causes a quantity of mayhem X that is high enough to prompt drastic action, but not so high that we can’t then recover. I say “would”, because COVID. Yeah we groped around for mitigation while it was happening, but in terms of lessons learned? It’s like there’s been a collective decision to focus on minutiae while ignoring the elephant in the room, and indeed feeding the elephant, idly fondling its trunk, bringing it lady elephants to make babies with (ok, the analogy is getting strained at this point…). So my mayhem value X would appear to significantly exceed “several million deaths.”
Right, you need a minor thing, easily interpretable as applying to everything, to happen first. For covid, that was the late 2019 outbreak in China + small # of rising cases locally - we could all see where things were headed.
But you can do that with AI too. You just have to run the simulations, and I think you can produce evidence (and promote it) that would be at least as compelling to enough relevant people.
But my focus is on the despair that at least some people in this thread are exuding. There are things that can be done. There are things that any of us, no matter how non-technical or powerless, can do to help, just like every other large coordination problem in history. There is no reason for despair.
That's why I say "accurate, extrapolatable-the-the-public evidence." With covid we ALSO coordinated shut down the world for varying durations on the basis of literally a chart showing a line going up to the right and some pictures from China.
So yes, I think my case that - if persuaded with sufficiently convincing evidence - we can coordinate massive global action: works just fine. Obviously yes, things did not go perfectly and we did things we shouldn't have, and we did not do things we should have. The % of those errors we have for future events will depend on how accurately we understand the threat, and it's likelihood and proximity - we are in the "before the next disaster" phase right now. So get out there and find your "upwards to the right" chart, if it exists.
I’m not disputing your case for action — far from it. “Very little hope” is not the same as “despair”, and arguably should be more motivating than “somewhat more hope.” Unfortunately, most of my remaining hope on this would come from a scenario where some less-than-all-powerful AGI-esque thing causes a quantity of mayhem X that is high enough to prompt drastic action, but not so high that we can’t then recover. I say “would”, because COVID. Yeah we groped around for mitigation while it was happening, but in terms of lessons learned? It’s like there’s been a collective decision to focus on minutiae while ignoring the elephant in the room, and indeed feeding the elephant, idly fondling its trunk, bringing it lady elephants to make babies with (ok, the analogy is getting strained at this point…). So my mayhem value X would appear to significantly exceed “several million deaths.”
Right, you need a minor thing, easily interpretable as applying to everything, to happen first. For covid, that was the late 2019 outbreak in China + small # of rising cases locally - we could all see where things were headed.
But you can do that with AI too. You just have to run the simulations, and I think you can produce evidence (and promote it) that would be at least as compelling to enough relevant people.
But my focus is on the despair that at least some people in this thread are exuding. There are things that can be done. There are things that any of us, no matter how non-technical or powerless, can do to help, just like every other large coordination problem in history. There is no reason for despair.