I’m an aspiring EA / rationalist. My previous posts were intended in a slightly tongue-in-cheek way (but conveying ideas I take seriously) - future posts will endeavour to lay out my ideas more clearly/ in keeping with LW norms.
Dzoldzaya
Karma: 108
KOL = Key Opinion Leaders, as in a small group of influential people within the neo-Luddite space. My argument here was simply that people concerned about AI alignment need to be politically astute, and more willing to find allies with whom they may be less aligned.
I think it’s probably a problem that those interested in AI alignment are far more aligned with techno-optimists, who I see as pretty dangerous allies, than more cautious, less technologically sophisticated groups, (bureaucrats or neo-Luddites).
Don’t know why you feel the need to use my unrelated post to attempt to discredit my comment here- strikes me as pretty bad form on your part. But, to state the obvious, a 40% shot at a desirable outcome is obviously not a call to action if the 60% is very undesirable (I mention that negative outcomes involve either extinction or worse).
Of course there are no such posts, and I hoped that people would read it in this spirit! I’m in fact arguing against that elephant in many a room where people are discussing collapse and x-risk.
I’m sure many people have thought:
a= x-risk this century, and b= chance of non-recovery post collapse and c= likelihood of future society reaching modernity at a stage when they’re better organised than current society at addressing the age of perils.
If a>b and c > 0.5, and we accept longtermism, then collapse seems desirable. If we add fairly pessimistic views about ongoing moral tragedies, or ideas like antinatalism and negative utilitarianism, it tips the balance further towards collapse.
This post is an expression of this dilemma- I feel it captures the tone I was hoping for… but no-one else seems to like it, unfortunately.
Definitely relevant, but not sure if it supports my argument that we shouldn’t try to induce collapse.
This post is about unilaterally taking a very radical action to avert a potentially worse scenario that inertia and race dynamics are pushing us towards. That looks like classic ‘unilateralist’s benediction’ territory.
Thanks! I’m actually a more serious EA type in everyday life, but my lesswrong alter ego is proudly Kakistocurious.
Sorry, I think the logic of my 60% figure was imprecise.
If we’re just talking about 1 person attempting to deliberately induce societal collapse, the chances of any kind of impact, either way, would be relatively low (depending on the person), so the 60% would seem a bit meaningless there.
If we’re talking about whether it’s worth developing a more serious movement to initiate societal collapse (potentially in an optimal way) or to plan for it as an option, I think there are arguments both ways, but I’d lean against it being a good idea because of the risks laid out in the post. This is what my 60% figure was aiming at.
If we’re talking about a world where we have successfully induced collapse (in a way that allows society to rebuild in some way), would this be a better or worse world, in expectation? This is the question I was really hinting at with this post, and I would definitely dispute your 5 in 1000 claim if this was the question you were thinking of.
If we’re Eliezer-level pessimistic about TAI timelines, serious about the horrors of factory farming (and perhaps antinatalism), and optimistic about moral progress in the absence of technological progress, I think this question gets very interesting.
Collapse Might Not Be Desirable
When considering whether to delay AI, the choice before us is not merely whether to accelerate or decelerate the technology. We can choose what type of regulations are adopted, and some options are much better than others.
We (the AI Safety community/ generally alignment-concerned people/ EAs) almost definitely can’t choose what type of regulations are adopted. If we’re very lucky/ dedicated we might be able to get a place at the table. Everyone else at the table will be members of slightly, or very, misaligned interest groups who we have to compromise with.
Various stripes of “Neo-Luddite” and AI-x-risk people have different concerns, but this is how political alliances work. You get at the table and work out what you have in common. We can try to take a leadership role in this alliance, with safety/ alignment as our bottom line- we’ll probably be a smaller interest group than the growing ranks of newly unemployed creatives, but we could be more professionalised and aware of how to enact political change.
If we could persuade an important neo-Luddite ‘KOL’ to share our concerns about x-risk and alignment, this could make them a really valuable ally. This isn’t too unrealistic- I suspect that, once you start feeling critical towards AI for taking your livelihood, it’s much easier to see it as an existential menace.
Adopting the wrong AI regulations could lock us into a suboptimal regime that may be difficult or impossible to leave. So we should likely be careful not [to] endorse a proposal because it’s “better than nothing” unless it’s also literally the only chance we get to delay AI.
Expecting anything close to optimal regulation in the current national/ international order on the first shot is surely folly. We should endorse any proposal that is “better than nothing” while factoring potential suboptimal regime shifts into our equations.
I’m not sure if it’s better or worse that the longtermist funder is the target of oncoming hate. I think it means that the ‘all those nice EAs stopping malaria and giving what they can’ narrative should remain positive, while the ‘crazy tech billionaire giving his money to stop an AI apocalypse’ narrative will be more negative, but quite distinct from the former.
If it were the other way around, and SBF had been the global poverty guy, that probably would have spawned an interesting societal reflection about consequentialism vs deontology. As it is, longtermism/ AI is too niche for it to be generally seen as altruistic rather than a techie pet project to most people, though.
I’d say that, in reality, open borders do require action by the state; just like deregulation, it’s something that is often portrayed as ‘inaction’, but actually ends up quite complex in practice. The EU/ Schengen zone, for example, needs to harmonise residency policies and external borders, develop cross-border policing and justice coordination between nations, harmonise workers’ rights and healthcare, all without a common language or a common defence force. Especially as most laws are designed for a world with borders, you would need to change a lot to implement open borders. Of course, you could just completely scrap the border police, visas, and any border checks and see what happens, but I doubt that anyone is seriously suggesting that.
I’m not arguing that we should always invoke a strict precautionary principle based on the status quo, although a mild ‘Chesterton’s Fence’ precautionary principle until we understand the facts on the ground is always prudent. Either way, any change we want to make or campaign for will be marginal, based on a country’s specific circumstances (imagine Canada vs. Israel). I presume you agree that it’s a bit ridiculous to have a hard border between the US and Canada, but I presume you wouldn’t recommend that Israel opens her gates freely to the Arab world.
As for the ‘policy paralysis’ idea, any policy I would suggest or campaign for would be at a far, far smaller scale than open borders. This is both for practical reasons, and out of a precautionary principle. I think it would probably be more ethical if the UK spent 2+% of its budget on foreign aid, for example, but I think that’s both politically impossible and could possibly have adverse consequences (if we cut other parts of the budget), but I’ve contributed to an (unsuccessful) campaign to keep the aid budget from going down from 0.7%, which I’m very confident is the ethically superior choice, and whose adverse consequences would be much smaller.
With migration, we all have a basic understanding of how migration can work and how it can cause harm, therefore any intervention I’d propose would try to harness these benefits and mitigate the harm based on this model. I might dedicate some time or effort to loosening restrictions concerning a certain population (I would be in favour of post-Brexit free movement to the UK by Aussies and Kiwis, for example), but I might also support restricting certain kinds of migration. I’ve read Caplan’s book and a lot of migration literature, and while I’m generally pro- migration at the margins, I don’t think we have anything like good evidence that open borders work, largely because all the evidence is theoretical and/ or based on controlled migration (or migration within a group of similar income countries etc.).
Great review, I agree very strongly with the America-centric nature of the book being annoying and misleading, but I have a few niggling points:
“Do more populous countries have greater growth in the long run? If so, this points us in the direction of open borders.” I think this is true, that populous countries have greater growth on the whole. But this doesn’t seem to point to open borders. I think there is a selection effect whereby only countries with a particularly strong system of governance don’t split up into other countries, and it’s the strong governance that creates growth. And some large nations don’t even have open borders between regions; it’s actually easier to move from rural Poland to Paris for work, and benefit from public services there, than it is to move from rural Hunan to Shanghai.
“There are various arguments related to longtermism that Caplan didn’t use. The downsides of immigration (higher crime, perhaps draining the government’s budget) are temporary but the upsides (higher economic growth) bear their fruit over centuries and will likely affect billions of future people.” I disagree with this, connected to reasons you mention later. The downsides can continue, even accelerate, for a good few generations, and then become fundamentally unpredictable; in France, for example, children of immigrants are more likely to commit crimes and be unemployed than 1st gen immigrants. Muslim migrants are less likely to want to assimilate to European countries than they used to be; a 2nd gen Muslim woman in Bradford is far more likely now to wear a Burqa than a 1st gen woman was 20 or 30 years ago, for example. For most people concerned about immigration, it’s the fact that their country (imagine Ukraine/ England/ Tibet/ Israel/ Luxembourg) won’t be their country anymore in a few decades (or centuries) that worries them. Tibetans have seen the number of Han Chinese in their nation rapidly increasing over the past 50 years, and they reasonably fear that once Tibet is 60-70% Han, they’ll either have to assimilate or be reduced to something akin to Aboriginals on reservations. Similarly, the theme of Soumission by Houellebecq is arguing something similar for France; if the rate of Islamic immigration continues to increase, then it’s possible that within a few decades an Islamist candidate could start implementing Islamic law within a European country (as we see regionally already, with divorce courts). You can only imagine if Israel were to allow open borders with her neighbours… This is to say that negative impacts of open borders may not be temporary, and could affect the mid-to-long-term identity, culture, norms and stability of a nation. It’s hard to make concrete predictions, but I’m tempted to predict that low-immigration, high-GDP countries (or countries with very selective migration, like Switzerland) will be more politically stable in the coming few decades. Paul Collier explains this dilemma well in his book, Exodus.
“The countries that are the closest to having open borders are the Gulf states; they have many migrant workers from countries like Bangladesh and Sri Lanka.” This seems the strangest line of the review. Depending on your definition, I’d say that the Gulf States have the opposite of open borders. They previously had targeted immigration policies allowing other Arabs from the MENA region to work there, then they caused too many social problems, so the Gulf countries threw them all out, and invited targeted immigration from a few specific poor countries (with people who didn’t speak Arabic, and therefore wouldn’t get involved in local politics). There is actually a decent amount of social mobility in these countries (a strangely high proportion of my friends are Gulf-state Indians based in Europe), so I wouldn’t be too worried about long-term racial segregation. If you see this story of poor South Asians taken out of poverty by working in the gulf, the Gulf States are a strong argument for a very ‘non-open borders’ way of doing mass immigration: inviting large numbers of migrants to come to a country on guest worker schemes, with very limited rights. Although these migrants have no social support, and can be thrown out on the whim of the recipient country, they can make loads of money compared to back home. I actually think this might be a really good idea, and this is also similar to what Chinese cities do with domestic migration, which avoids parts of Beijing and Shanghai turning into huge shantytowns. However, as you mention, having a ‘second-class’ population sits poorly with European norms and sensibilities.
I don’t think your claim that ‘the intervention is closed borders’ is quite right. The status quo is differing levels of restriction of movement between regions and nations that have developed as a result of long historical processes. As the EU demonstrates, it can be as difficult to manage (relatively) open borders as it is to manage (relatively) closed borders. The status quo is that countries and regions choose different levels of migration, and different kinds of migration, to suit their values and needs, therefore the UK, the EU, the US, Canada, China, the Gulf Countries and Australia have all developed hugely different migration policies, while all remaining largely successful as nations over the last few decades.
Arguing for ’100% closed vs. 100% open borders’ gets very abstract and it’s difficult to argue from the facts rather than ideology, because fully open or closed borders are so rare. Although it can be useful to try and nudge the Overton window, the intervention is always going to be a marginal shift in the status quo, but even then, we should be humble about the downsides of marginal shifts either way.
Well, inducing a mass societal collapse is perhaps one of the few ways that a small group of people with no political power or allies would be able to significantly influence AI policy. But, as I stressed in my post, that is probably a bad idea, so you shouldn’t do it.