Someone should commission new moral weights work in the next year

The moral weights of animals seems like one of the most important inputs into cause prioritization. The difference between whether we use RP weights or neuron count is the difference between whether the present contains more happiness than suffering, and potentially whether humanity has been overall good or bad for wellbeing. 

This also poses challenges to the future. Averting catastrophes is profoundly insufficient if the default trajectory for wellbeing is negative (and potentially worsening). Indeed, if the default trajectory is negative (and we have no good ways of changing it) we can imagine the universe giving a sigh of relief if we were filtered out of the cosmic pool of awareness. 

Given the profound importance for cause prioritization -- if the present is overall negative for wellbeing I think it implies we should focus much, much more on making the future and present go well than go long -- we should have several independent well resourced attempts to answer the question of "how do we weigh the wellbeing of animals versus humans?" 

There's also the historical fact that the last time someone (the Imperial Japanese) tried this on the U.S. (Pearl Harbor), it ended up being disastrous for them.

I think there's an important disanology with a first strike against the USA now. Japan woefully underestimated the USA's manufacturing ability. We rebuilt all our lost fleet and then some. I really doubt we'd be able to rebuild in the time it'd become relevant to actually challenge the PRC -- indeed you seem to be aware of this issue. I think it's better to imagine China as the USA during WW2 in terms of capacity to fight an industrial war. 

And yes, I agree that trying to match China industrially is setting yourself up for failure. I think there could be a smarter way as you note (and note I noted! - Hah!) by focusing on less flashy but highly lethal and asymmetrical systems. If I was China I'd really want to know how well Taiwan's autonomous naval drones work. 

In general regarding EA involvement in this though, I think a lot of thought has already been put into these concerns by the people within the NatSec establishment, such that I'm not sure what EAs can actually add to the equation.

Yep, I think this is super fair. I'm not sure if we have something to contribute, but I lean towards an optimistic maybe. One of the key points I wanted to make is that given the risks involved, it seems like we should try a little bit harder to see if we can carefully contribute in a clearly beneficial way before throwing in the towel. 

Also thanks for the kind words! I've found your comments very helpful. You may be right! I think we should try harder to find out, the stakes are high.

I don't think China would need to be deterred forever. I think it would need to probably be deterred for the next ~10 years. After which... 

• India will probably step into global power status, and become a peer competitor with China. Unless they've settled their substantial border issues, this seems like a very plausible source of conflict from the PRC's perspective. Most interstate conflicts are border skirmishes (I think).
• I note this in the essay, but the PRC is in a rough neighborhood. If they act on Taiwan they need to be confident they can also ensure that a competitor like India doesn't take the opportunity to make a move to resolve a border dispute. The same goes for any adversary who they have conflicting territorial claims with in the South China sea.
• Whatever is happening with AI will probably pan out and for better or worse we're in a world beyond our ability to easily imagine.
• Demographic issues in China will start to bite which I think will relate to a broader risk aversion (a pretty weak reason). 

More military power could also increase the costs in case of an invasion, so you need to be quite confident that deterrence will work and continue working indefinitely.

Yes, I agree on the first part. Not on the "indefinitely" part. 

To me, the most tractable part of this equation seems to be the chance that the USA intervenes, but I'm profoundly uncertain if it would be better for the world if this were higher or lower. Do you have any views on that?

Reducing the chance the USA intervenes doesn't strike me as very tractable. What are you imagining here? 

• Grassroots campaigns towards isolationism (isolationism is now associated with the right so not sure about that)
• General advocacy of austerity with cuts to defense budget (that could do it but less direct and very hard to do).
• Make some deal with China conditional on non-intervention in Taiwan?  

I'm a little nervous around geopolitics as a philanthropic cause area cause it seems so easy to harm when information and calculations are imperfect (and our information will be extremely imperfect). [...]

Do you think there are clear things we can do in this space to mitigate downside risks?

I am also nervous and agree with the downside risks. I suspect there are some clear actions we can take, but I think we definitely need more research to know which are the best of the bunch. 

I do stand by one of my main arguments that encouraging Taiwan to defend itself more smartly will make it less likely China invades. But I'm not committed to this as being the most cost-effective option. There are many other options I spent less time on like advocating that the USA minimize unhelpful and inflammatory diplomatic showmanship like the Pelosi visit.

You're welcome! Thanks for the kind words. 

I think the likelihood of a blockade is lower than the metaculus community, but this is a low volume question. 

I disagree with Hugh White. This is for a couple weak reasons. 

• Many of the experts I trust seem to think this is a less likely scenario. I don't know Hugh White well, but he came off a bit more "talking head" than calibrated analyst (or hedgehog than fox) to me (I'll probably try and explain this in another comment).
• My model of Xi / the PRC is that they want to minimize the time they're exposed to potential diplomatic pushback. I guess that fait accompli = good, with the example of the Russian occupation of Crimea as ideal. Act before anyone can really get angry. A blockade seems like something they may want to threaten but I doubt they want to engage in.
  • It would potentially take a really long time to squeeze Taiwan (6 months to 2 years?). All this time could potentially galvanize global support for Taiwan, and consolidate anti-PRC / resistance sentiment within the country. It's hard to justify internationally that you are trying to starve a whole country. It's probably harder to justify than what I expect the PRC would prefer which is to actually do what Russia planned on doing -- a quick 3 day operation where they take out military sites and the head of government and say "we had information Taiwan was about to declare independence. We have taken out the separatists, and protected the unity of the Chinese people."  
  • During this time it wouldn't be too hard for Taiwan to ensure their fabs are destroyed if the PRC invades and to smuggle out key TSMC staff to Japan / USA in return for USA airlifts or something.
  • Indeed, I'm pretty sure blockades are considered an act of war under international law. If so, this would potentially allow Taiwan / the USA to attempt to break the blockade at a time of their choosing with justification. So I think a blockade could potentially leave the Chinese navy very vulnerable.
  • Blockades don't have a strong track record of forcing capitulation. See Israel and Gaza, USA and Cuba, or USSR and Berlin.  
  • If they fail then the PRC has lost their opportunity for a surprise attack and a good amount of credibility.
• I strongly doubt the USA would sabotage Taiwan fabs. I mean, I think if anything Taiwan would do this themselves (but only if a Chinese invasion was inevitable). Sure the USA, may want to make sure Taiwan has these plans, but I don't think they'd unilaterally sabotage the fabs. Is there any precedent for anything like this? We didn't even destroy a fraction our massive amounts of valuable US hardware from getting into the hands of the Taliban -- granted those were tactical rather than strategic assets. Maybe I'd update this if it turns out the USA was behind Nordstream 2 -- I thought that turned out to be Ukraine though. 

But I'd welcome more forecasts and breakdowns of these scenarios! 

I would love to hear someone make a more detailed argument about how this would affect things. 

• Naively, invasion --> Taiwan fabs destroyed --> timelines pushed back a couple of years.
• Longer timelines typically considered good because it means more time to do safety work.
• But my gut says that when the supply chain is rebuilt this would mean China has a relatively larger share of high end chips (but very, very unclear) --> race dynamic is tighter --> world more dangerous.
• Net effect --> ???

I guess my question is: how much risk reduction would there be if USA stopped messing with the world?

If the messing around pertains to Taiwan --  a naive implication of my model is if the USA had a 0% chance of intervening in Taiwan this would decrease the risk of catastrophe by around 1% (percentage point). This seems difficult to influence though. 

Howdy Huw, 

I think that Xi has shown himself less sensitive to economic pressure than previous premiers or the leaders of more liberal countries (see zero Covid). I'm concerned that he sees himself in some manner similar to Putin -- in that they both have historical missions to fulfill. The great rejuvenation of the Chinese nation in Xi's case. A lot of his action's and PRC industrial policy (squeezing certain seemingly decadent sectors like gaming and propping up manufacturing) is probably bad for growth in the long run. But they are consistent with a China that bases its economy on something more material and connected to power (and the ability to win a protracted war). 

This is all to say, I don't see trade as a major detterent even if it was a possible threat. But I think when it comes to trade, I'm concerned that stable authoritarian regimes just have more leverage due to their willingness to replace consumption with patriotism in a way that seems to have no equivalent in modern democracies. 

There's still potentially something here, I just haven't given it too must time or thought. One of the many good questions I wish we had better answers to. 

The timeline would depend on the magnitude of the scale. I think a good analogy is to see how long it's taking Poland's military to absorb it's sizable surge in investment spending. In short I'd guess anything substantial would take years. 

And definitely it's a risk. I think a key assumption is that the PLA isn't yet ready to invade. If we were already in that window, then I'd be more concerned by the risks of that strategy. 

I'm sure there's some good empirical analysis that is either possible (or more likely) already been done that could inform this question -- but it's just far enough outside my wheelhouse that I haven't found it. 

Are there any EA organisations that fund creation of popular media in order to deliver impact? If not, why not?

The closest thing I've seen is https://www.shouldwestudio.com . The founder was on 80k and Clearer Thinking recently. 

This is a good answer and gives a lot of important context, thanks Joseph. 

Regarding the conscription, I tried but failed to get a sense if it got more serious since they went from 4 months to a full year of service.