This would absolutely be a bad message to use, voters don't care about aid at all. You'd just use the best message-tested stuff available which generically moves the needle in the direction you want it to go

I helped work on this piece along with some other research attempting to assess tractability. I think it wasn't super obvious that it was the best way to spend money but it was probably cost-competitive with many top donation opportunities in expectation. There also may have been ways we could have influenced things early on if we'd been putting in the effort in, say, 2023 (e.g. trying to get Biden out faster). 

Politics work is basically always going to very low probability, high reward. In this election I'd argue the expected impacts counterbalanced the low probability of success. 

It's true that some EAs work in government but I think this piece lays out pretty well what that actually looks like, and it doesn't typically involve politics - it's more civil service type things. I'm pretty sure I know most EAs who work on direct pollitical work (e.g. elections) and it's quite a small number. 

That said, yeah, it's good that there are some people working in government and that does help broader EA understand the political situation a little better. 

Agree this is a very thorny problem and I am unsure of how to deal with it. Suspect there's some degree to which you can balance it usefully and that it's worth paying some cost of looking partisan but ultimately it's not really viable to coordinate everyone's level of partisanship. 

I think a big part of mitigating the costs is just trying to avoid the sense that you're speaking on behalf of EA or AI safety when talking about partisan stuff

I made a prediction that foreign aid would be cut significantly in ~September of last year (see below) so seems like there's some degree to which at least some cuts were predictable. I think the intervention I advocated for at the time,  stopping Trump from being elected, would have been the most straightforward action to take (and I think EV of doing that from even a pure global health perspective looked alright). 

I didn't predict specific DOGE cuts. That said, if one had, then trying to get the message to Musk that this matters could have been a reasonable action to take (and would have been usefully informed by better political analysis). Plausibly there's some messaging stuff one could do? 

Otherwise, the best thing to do might have been to have some contingency planning for large aid cuts? I'm not sure how much counterfactual value having those plans would end up being but it seems possible that with some preparation beforehand one could keep larger parts of PEPFAR alive maybe? Certainly it seemed like the sector was really overwhelmed upon learning of USAID cuts and that seems like some indication that more preparation would have been useful. 

Overall, I don't feel too strongly that knowing DOGE cuts were coming would have been super high leverage but I think it exemplifies politics as being something which can have extremely far reaching impacts on cause areas that EAs care about - even far beyond anything happening internal to the field. Same thing seems true of AI as well as animal welfare and pandemic prevention. In the extreme, the end of democracy in the US would seem pretty likely to have a bigger negative impact (by far) on all EA cause areas than basically anything one could do internal to the cause area. 

Exact prediction about aid cuts (which I made after maybe 1-2 hours of looking into this): 
"If [Harris] spends at the same level as Biden (and Trump reverts to his prior spending), getting her into office would lead to ~$16 billion going to international aid [over the full term] that otherwise wouldn’t have. " 
 

Thank you! Fixed

That's my bad, I did say 'automated' and should have been 'automatable'. Have now corrected to clarify

Do you have anything you recommend reading on that?

I guess I see a lot of the value of people at labs happening around the time of AGI and in the period leading up to ASI (if we get there). At that point I expect things to be very locked down such that external researchers don't really know what's happening and have a tough time interacting with lab insiders. I thought this recent post from you kind of supported the claim that working inside the labs would be good? - i.e. surely 11 people on the inside is better than 10? (and 30 far far better)

I do agree OS models help with all this and I guess it's true that we kinda know the architecture and maybe internal models won't diverge in any fundamental way from what's available OS. To the extent OS keeps going warning shots do seem more likely - I guess it'll be pretty decisive if the PRC lets Deepseek keep OSing their stuff (I kinda suspect not? But no idea really). 

I guess rather than concrete implications I should indicate these are more 'updates given more internal deployment' some of which are pushed back against by surprisingly capable OS models (maybe I'll add some caveats)

I think it's very reasonable to say that 2008 and 2012 were unusual. Obama is widely recognized as a generational political talent among those in Dem politics. People seem to look back on, especially 2008, as a game-changing election year with really impressive work by the Obama team. This could be rationalization of what were effectively normal margins of victory (assuming this model is correct) but I think it matches the comparative vibes pretty well at the time vs now. 

As for changes over the past 20+ years, I think it's reasonable to say that there's been fundamental shifts since the 90s:

• Polarization has increased a lot 
• The analytical and moneyball nature of campaigns has increased by a ton. Campaigns now know far more about what's happening on the ground, how much adversaries spend, and what works.
• Trump is a highly unusual figure which seems likely to lead to some divergence
• The internet & good targeting have become major things 

Agree that 5-10% probability isn't cause for rejection of the hypothesis but given we're working with 6 data points, I think it should be cause for suspicion. I wouldn't put a ton of weight on this but 5% is at the level of statistical significance so it seems reasonable to tentatively reject that formulation of the model.

Trump vs Biden favorability was +3 for Trump in 2020, Obama was +7 on McCain around election day (average likely >7 points in Sept/Oct 2008). Kamala is +3 vs Trump today. So that's some indication of when things are close. Couldn't quickly find this for the 2000 election.

I think this is all very reasonable and I have been working under the assumption of one votes in PA leading to a 1 in 2 million chance of flipping the election. That said, I think this might be too conservative, potentially by a lot (and maybe I need to update my estimate).  

Of the past 6 elections 3 were exceedingly close. Probably in the 95th percentile (for 2016 & 2020) and 99.99th percentile (for 2000) for models based off polling alone. For 2020 this was even the case when the popular vote for Biden was +8-10 points all year (so maybe that one would also have been a 99th percentile result?). Seems like if the model performs this badly it may be missing something crucial (or it's just a coincidental series of outliers). 

I don't really understand the underlying dynamics and don't have a good guess as to what mechanisms might explain them. However, it seems to suggest that maybe extrapolating purely from polling data is insufficient and there's some background processes that lead to much tighter elections than one might expect. 

Some incredibly rough guesses for mechanisms that could be at play here (I suspect these are mostly wrong but maybe have something to them):

• Something something polarization, steady voting blocs for Rep & Dem aren't shifting much year to year. This means we should expect similar margins this year as 2016 & 2020. 
• Some balancing out process where politicians are adjusting their platform, messaging, etc to react to their adversary and this ends up increasing how close elections get.
• Maybe something where voters have local information on whether the person they don't like is more likely to win and they then feel more motivated to vote? Turns out, in aggregate, this local information is pretty accurate and leads to tighter-than-expected elections.
• Maybe political parties/donors observe how much their adversary spends in a given state and are consistently able to spend to counteract their efforts. This maybe provides a balancing effect that tightens the race. This would have the unfortunate consequence that visible spending is much less effective - but maybe implies that smaller, more under-the-radar, projects are better.