Will the world's elites navigate the creation of AI just fine?

lukeprog

One open question in AI risk strategy is: Can we trust the world's elite decision-makers (hereafter "elites") to navigate the creation of human-level AI (and beyond) just fine, without the kinds of special efforts that e.g. Bostrom and Yudkowsky think are needed?

Some reasons for concern include:

Otherwise smart people say unreasonable things about AI safety.
Many people who believed AI was around the corner didn't take safety very seriously.
Elites have failed to navigate many important issues wisely (2008 financial crisis, climate change, Iraq War, etc.), for a variety of reasons.
AI may arrive rather suddenly, leaving little time for preparation.

But if you were trying to argue for hope, you might argue along these lines (presented for the sake of argument; I don't actually endorse this argument):

If AI is preceded by visible signals, elites are likely to take safety measures. Effective measures were taken to address asteroid risk. Large resources are devoted to mitigating climate change risks. Personal and tribal selfishness align with AI risk-reduction in a way they may not align on climate change. Availability of information is increasing over time.
AI is likely to be preceded by visible signals. Conceptual insights often take years of incremental tweaking. In vision, speech, games, compression, robotics, and other fields, performance curves are mostly smooth. "Human-level performance at X" benchmarks influence perceptions and should be more exhaustive and come more rapidly as AI approaches. Recursive self-improvement capabilities could be charted, and are likely to be AI-complete. If AI succeeds, it will likely succeed for reasons comprehensible by the AI researchers of the time.
Therefore, safety measures will likely be taken.
If safety measures are taken, then elites will navigate the creation of AI just fine. Corporate and government leaders can use simple heuristics (e.g. Nobel prizes) to access the upper end of expert opinion. AI designs with easily tailored tendency to act may be the easiest to build. The use of early AIs to solve AI safety problems creates an attractor for "safe, powerful AI." Arms races not insurmountable.

The basic structure of this 'argument for hope' is due to Carl Shulman, though he doesn't necessarily endorse the details. (Also, it's just a rough argument, and as stated is not deductively valid.)

Personally, I am not very comforted by this argument because:

Elites often fail to take effective action despite plenty of warning.
I think there's a >10% chance AI will not be preceded by visible signals.
I think the elites' safety measures will likely be insufficient.

Obviously, there's a lot more for me to spell out here, and some of it may be unclear. The reason I'm posting these thoughts in such a rough state is so that MIRI can get some help on our research into this question.

In particular, I'd like to know:

Which historical events are analogous to AI risk in some important ways? Possibilities include: nuclear weapons, climate change, recombinant DNA, nanotechnology, chloroflourocarbons, asteroids, cyberterrorism, Spanish flu, the 2008 financial crisis, and large wars.
What are some good resources (e.g. books) for investigating the relevance of these analogies to AI risk (for the purposes of illuminating elites' likely response to AI risk)?
What are some good studies on elites' decision-making abilities in general?
Has the increasing availability of information in the past century noticeably improved elite decision-making?

Some reasons for concern include:

Otherwise smart people say unreasonable things about AI safety.
Many people who believed AI was around the corner didn't take safety very seriously.
Elites have failed to navigate many important issues wisely (2008 financial crisis, climate change, Iraq War, etc.), for a variety of reasons.
AI may arrive rather suddenly, leaving little time for preparation.

But if you were trying to argue for hope, you might argue along these lines (presented for the sake of argument; I don't actually endorse this argument):

If AI is preceded by visible signals, elites are likely to take safety measures. Effective measures were taken to address asteroid risk. Large resources are devoted to mitigating climate change risks. Personal and tribal selfishness align with AI risk-reduction in a way they may not align on climate change. Availability of information is increasing over time.
AI is likely to be preceded by visible signals. Conceptual insights often take years of incremental tweaking. In vision, speech, games, compression, robotics, and other fields, performance curves are mostly smooth. "Human-level performance at X" benchmarks influence perceptions and should be more exhaustive and come more rapidly as AI approaches. Recursive self-improvement capabilities could be charted, and are likely to be AI-complete. If AI succeeds, it will likely succeed for reasons comprehensible by the AI researchers of the time.
Therefore, safety measures will likely be taken.
If safety measures are taken, then elites will navigate the creation of AI just fine. Corporate and government leaders can use simple heuristics (e.g. Nobel prizes) to access the upper end of expert opinion. AI designs with easily tailored tendency to act may be the easiest to build. The use of early AIs to solve AI safety problems creates an attractor for "safe, powerful AI." Arms races not insurmountable.

The basic structure of this 'argument for hope' is due to Carl Shulman, though he doesn't necessarily endorse the details. (Also, it's just a rough argument, and as stated is not deductively valid.)

Personally, I am not very comforted by this argument because:

Elites often fail to take effective action despite plenty of warning.
I think there's a >10% chance AI will not be preceded by visible signals.
I think the elites' safety measures will likely be insufficient.

In particular, I'd like to know:

Which historical events are analogous to AI risk in some important ways? Possibilities include: nuclear weapons, climate change, recombinant DNA, nanotechnology, chloroflourocarbons, asteroids, cyberterrorism, Spanish flu, the 2008 financial crisis, and large wars.
What are some good resources (e.g. books) for investigating the relevance of these analogies to AI risk (for the purposes of illuminating elites' likely response to AI risk)?
What are some good studies on elites' decision-making abilities in general?
Has the increasing availability of information in the past century noticeably improved elite decision-making?

Thanks for engaging.

Why would a good AI policy be one which takes as a model a universe where world destroying weapons in the hands of incredibly unstable governments controlled by glorified tribal chieftains is not that bad of a situation? Almost but not quite destroying ourselves does not reflect well on our abilities. The Cold War as a good example of averting bad outcomes? Eh.

The point is that I would have expected things to be worse, and that I imagine that a lot of others would have as well.

This is assuming that people understand what makes an AI so dangerous - calling an AI a global catastrophic risk isn't going to motivate anyone who thinks you can just unplug the thing (and even worse if it does motivate them, since then you have someone who is running around thinking the AI problem is trivial).

I think that people will understand what makes AI dangerous. The arguments aren't difficult to understand.

The fact that someone is powerful is evidence that they are good at gaining a reputation in their specific field, but I don't see how this is evidence for rationality as such (and if we are redefining it to include dictators and crony politicians, I don't know what to say),

Broadly, the most powerful countries are the ones with the most rational leadership (where here I mean "rational with respect to being able to run a country," which is relevant), and I expect this trend to continue.

Also, wealth is skewing toward more rational people over time, and wealthy people have political bargaining power.

why would someone who has no experience with these kind of issues suddenly grab it out of the space of all possible ideas he could possibly be thinking about?

Political leaders have policy advisors, and policy advisors listen to scientists. I expect that AI safety issues will percolate through the scientific community before long.

It seems like you are claiming that AI safety does not require a substantial shift in perspective (I'm taking this as the reason why you are optimistic, since my cynicism tells me that expecting a drastic shift is a rather improbable event) - rather, we can just keep chugging along because nice things can be "expected to increase over time", and this somehow will result in the kind of society we need. [...]

I agree that AI safety requires a substantial shift in perspective — what I'm claiming is that this change in perspective will occur organically substantially before the creation of AI is imminent.

Also I really don't know where you got that last idea - I can't imagine that most people would find AI safety more glamorous then, you know, actually building a robot.

You don't need "most people" to work on AI safety. It might suffice for 10% or fewer of the people who are working on AI to work on safety. There are lots of people who like to be big fish in a small pond, and this will motivate some AI researchers to work on safety even if safety isn't the most prestigious field.

If political leaders are sufficiently rational (as I expect them to be), they'll give research grants and prestige to people who work on AI safety.

I think that people will understand what makes AI dangerous. The arguments aren't difficult to understand.

We still get people occasionally who argue the point while reading through the Sequences, and that's a heavily filtered audience to begin with.

3wubbles13y

Things were a lot worse then everyone knew: Russia almost invaded Yugoslavia, which would have triggered a war according to newly declassified NSA journals, in the 1950's. The Cuban Missile Crisis could easily have gone hot, and several times early warning systems were triggered by accident. Of course, estimating what could have happened is quite hard.

36

Will the world's elites navigate the creation of AI just fine?

36

36

36

Will the world's elites navigate the creation of AI just fine?

36

36