I think different kinds of risks have different “distributions” of how much damage they do. For example, the majority of car crashes causes no injuries (but damage to the cars), a smaller number causes injuries and some causes fatalities, and the worst ones can cause multiple fatalities.
For other risks like structural failures (of buildings, dams, etc.) the distribution has a longer tail: in the worst case very many people can die. But the distribution still tapers off towards greater number of fatalities, and people sort have have a good idea of how bad it can get before the worst version happens.
For risks like war, the distribution has an even longer tail, and people are often caught by surprise how bad they can get.
But for AI risk, the distribution of damage caused is very weird. You have one distribution for AI causing harm due to its lack of common sense, where it might harm a few people, or possibly cause one death. Yet you have another distribution for AI taking over the world, with a high probability of killing everyone, a high probability of failing (and doing zero damage), and only a tiny bit of probability in between.
It’s very very hard to learn from experience in this case. Even the biggest wars tend to surprise everyone (despite having a relatively more predictable distribution).
That’s a cool way to frame damage risks, but I think your distribution for AI damage is for ASI, not AGI. I think it’s very reasonable that an AGI-based system may cause the type of damage that I am talking about.
Even if you believe that as soon as we achieve AGI, we’ll accelerate to ASI because AGI by definition is self-improving, it still takes time to train a model, and research is slow. I hope that the window b/w AGI & ASI is large enough for such a “Hiroshima event” to occur, so humanity wakes up to the risks of maligned AI systems.
PS: Sorry for the late response, I was offline for a couple of days
No need to say sorry for that! On a forum, there is no expectation to receive a reply. If every reply obligated the recipient to make another reply, comment chains will drag on forever.
You can freely wait a year before replying.
I’m worried that once a “Hiroshima event” occurs, humanity won’t have another chance. If the damage is caused by the AGI/ASI taking over places, then the more power it obtains, the more it can obtain even more power, so it won’t stop at any scale.
If the damage is caused by bad actors using an AGI to invent a very deadly technology, there is a decent chance humanity can survive, but it’s very uncertain. A technology can never be uninvented, and more and more people will know about it.
I think different kinds of risks have different “distributions” of how much damage they do. For example, the majority of car crashes causes no injuries (but damage to the cars), a smaller number causes injuries and some causes fatalities, and the worst ones can cause multiple fatalities.
For other risks like structural failures (of buildings, dams, etc.) the distribution has a longer tail: in the worst case very many people can die. But the distribution still tapers off towards greater number of fatalities, and people sort have have a good idea of how bad it can get before the worst version happens.
For risks like war, the distribution has an even longer tail, and people are often caught by surprise how bad they can get.
But for AI risk, the distribution of damage caused is very weird. You have one distribution for AI causing harm due to its lack of common sense, where it might harm a few people, or possibly cause one death. Yet you have another distribution for AI taking over the world, with a high probability of killing everyone, a high probability of failing (and doing zero damage), and only a tiny bit of probability in between.
It’s very very hard to learn from experience in this case. Even the biggest wars tend to surprise everyone (despite having a relatively more predictable distribution).
That’s a cool way to frame damage risks, but I think your distribution for AI damage is for ASI, not AGI. I think it’s very reasonable that an AGI-based system may cause the type of damage that I am talking about.
Even if you believe that as soon as we achieve AGI, we’ll accelerate to ASI because AGI by definition is self-improving, it still takes time to train a model, and research is slow. I hope that the window b/w AGI & ASI is large enough for such a “Hiroshima event” to occur, so humanity wakes up to the risks of maligned AI systems.
PS: Sorry for the late response, I was offline for a couple of days
No need to say sorry for that! On a forum, there is no expectation to receive a reply. If every reply obligated the recipient to make another reply, comment chains will drag on forever.
You can freely wait a year before replying.
I’m worried that once a “Hiroshima event” occurs, humanity won’t have another chance. If the damage is caused by the AGI/ASI taking over places, then the more power it obtains, the more it can obtain even more power, so it won’t stop at any scale.
If the damage is caused by bad actors using an AGI to invent a very deadly technology, there is a decent chance humanity can survive, but it’s very uncertain. A technology can never be uninvented, and more and more people will know about it.
Or more! (I was delighted to receive this reply.)