This will be resolved based on my judgement of the vibes of top Rationalist voices in 4 years.
If through their Tweets and Substack posts, I get the sense that they are happy with Trump's reforms, and that what got done was extremely impactful, vastly outweighs any harms, and was more sizable than what other post WWII presidents have accomplished, then I will resolve YES. Otherwise, NO.
EDIT: This would be judged from 1960 with JFK (68 years before 2028).
Update 2025-02-05 (PST) (AI summary of creator comment): Majority Requirement Clarification:
A majority of prominent rationalists must hold the view that Trump’s second term was the most positively impactful since 1960.
The final judgment will consider the vibes of rationalists across a spectrum, where the intensity of their views is weighed—if the core (top) voices are more fervent, they can outweigh a larger number of mildly opposing voices.
The process involves assessing gradations among rationalists, meaning that the consensus reflects the overall tendency rather than a strict headcount.
1,000Item 24 in today's ACX presumably bumps the probability here up slightly. (The rest of that post, not so much, I think.)
@dreev I'm actually surprised how against Doge Scott is. In my mind, it's status quo bias, as well as different priors about how bad government was before. I will admit I didn't realize Pepfar was still in limbo.
Lastly, one more crux could be from my experience as a founder: the bias toward action and shaking things up usually is a profound good and leads to very high expected results, whereas to most people it looks like flailing and destruction. I think the best managed orgs will always have a disruptive current and the government has had almost none for forever. See also my Mad Scientist theory.
@dreev If we're ignoring AI, nothing compares to PEPFAR, frankly. It's as efficient as GiveWell, has saved 19 million, and 6 million will die if it doesn't get out of limbo.
@JamesGrugett The problem is that government is not a tech company. You're trying to apply a heuristic from your own experience in a very weird way - if anything, historically speaking, you want governments to be relatively stable democracies with boring centrist-ish leaders rather than schizoposting dictator-aspiring people trying to remove checks and balances.
I think viewing government in an absolute "how bad is it compared to my ideal system" is a mistake. View government relative to other governments. There was corruption in the US government, but it's not nearly at the point where I'd throw it all away for a wildcard with bad economic policy who's worse on even free speech and expression than his opponent.
As for DOGE/Elon specifically, the problem is that it clearly fails cost/benefit.
Benefit:
- DOGE has saved about $3 billion ($10 per American). (Most of DOGE's other savings are things that were already spent and counted anyway, contracts not fully terminated, or firing workers who just re-enter the workforce, at best in a slightly more productive position.) It looks like DOGE's cuts are unlikely to break even with increase of federal spending due to inflation, according to the Kalshi market "how much spending will Trump and Elon cut?"
- Some DEI bloat removed.
- Some woke science (~2% of all grants under the Biden admin) removed.
Cost:
- Many important workers for agencies like NOAA, FAA fired.
- The CPFB has been shut down. The CPFB saved consumers $20 billion, so if you count that as straightforward savings, DOGE has actually lost $17 billion. Some might be phantom savings, because it isn't revenue for the federal government, it's just being moved from a firm to an individual, but overall I think CPFB was good enough that it's better than all the money DOGE saved.
- Some actual longstanding science grants and funding for science in universities removed. Some of which is important research that could save lives. I don't have estimates off the top of my head but I think Scott tried and concluded that it was pretty bad.
- Blatantly illegal actions that undermine the rule of law.
- Elon publicly calling to impeach judges who go against him, also undermining the rule of law.
- Other good websites, agencies, or programs that costed relatively little becoming defunct for seemingly no reason.
Even if you could say DOGE itself is net positive (which it clearly isn't), it pales in comparison to the tax cuts that Trump is making, which will increase the deficit.
But again, PEPFAR + actions that undermine democracy globally outweigh all of this. The only way I could see a turnaround at this point is if he agrees to fund most of the USAID programs under the State Department and implements some seriously good deregulatory policy. But considering how he just saved a corrupt union from automation, that's not a guarantee.
@JamesGrugett It’s worth noting that a lot of DOGE’s targets have been the places where the federal government was innovating the most. For example, they’re tearing out almost everything that the US Digital Service did, which was bringing the rest of the government to have modern websites that are easy to use. They’re tearing out electric car chargers. They’re removing the IRS’s free and easy filing service that was being developed and was a huge step forward.
This isn’t a “bias toward action”, it’s intentional destruction that is getting rid of the innovative parts of the government. The literal actual goal is to make the government more lethargic and less able to take action quickly.
In addition, like zephyr said, a government generally shouldn’t be chaotic. So much of what gets done doesn’t need innovation, it needs to be done consistently and without chaos. I don’t want my weather reports to not happen some days because they’re mixing things up. I want the dollar bills to be printed and not lost. I want my mail to be delivered on time.
And finally, PEPFAR on its own outweighs the benefits of any cost cutting efficiency that may happen (of which there really isn’t much). PEPFAR is more efficient than everything that EA does. If you had me choose between PEPFAR and the entire EA movement, I’d choose PEPFAR.
Given the amount of damage Trump has done in the first month of his administration and shows no signs of stopping, this seems very unlikely. My remaining sources of uncertainty are that:
1) The market creator is the largest YES holder. I generally trust James, but there is some subjectivity in the resolution giving room for motivated cognition.
2) Given Trump's disregard for liberal norms, there's some chance that by the time of resolution, prominent rationalists will be coerced to judge Trump positively in their public statements.
Regular reminder to go bet on the actual serious market that is not predicated on sheer delusion: https://manifold.markets/Balasar/will-prominent-rationalists-judge-t-Zg29U08R8g
https://www.lesswrong.com/posts/qYPHryHTNiJ2y6Fhi/the-paris-ai-anti-safety-summit
Trump has already been a disaster for AI Safety, per Zvi
The Trump Administration has made its position very clear. It intends not only to not prevent, but to hasten along and make more likely our collective annihilation. Hopes for international coordination to mitigate existential risks are utterly collapsing.
@Gabrielle yeah Trump is pretty much the worst nightmare for AI pause types.
Well, I expect it to get worse (government funding big computers). We'll see what happens after the 180 days mentioned in Trump's AI executive order has passed.
@jim Worth noting it is dream come true for the "AI safety is pseudoscience" types.
This one sounds like a slightly positive update:
https://www.astralcodexten.com/p/1daysooners-trump-ii-health-policy
Temper it by @ScottAlexander's mention of a ~10% chance of all the good things in that list happening. And of course the whole list is just for the domain of health policy. But we're looking at something short of fully unmitigated badness, I guess?
Surprise surprise, Scott Alexander isn't a fan of cancelling PEPFAR.
https://www.astralcodexten.com/p/money-saved-by-canceling-programs
Just to be clear, is this question specifically tied to things that Trump does, on purpose-ish, in office, before 2029—or does Trump get credited for things that happen as a consequence of his election and behavior?
If Trump resigns in 2025 after a disastrous run and Vance goes on to reform the presidency and have the most positively impactful term before 2029—is that a YES for Trump’s positive impact, in hindsight?
Or if Trump starts a war with Russia and China, causing all three to collapse and leading to the formation of the one-world-government with UBI and safe AGI for all…—before 2029 and in opposition to his personal wishes—positively impactful or NO?
(Or maybe his sudden and perhaps unintentional destruction of the global economy in 2025 delays the creation of ASI to 2028, immediately after the discovery of the three laws that really make safe AIs, but the super-benevolent ASI then removes Trump from his position as forever-president against his wishes?)
TLDR: is this a consequentialist reading of Trump’s second term?
@TannerNewell This is similar to the AGI question from yesterday. Suppose we agree that superintelligence has a 10% chance of killing everyone. And suppose Trump is pivotal in humanity's decision to roll the dice on that. If it works out, that's a massive positive impact, ex post. But the EV was wildly negative.
I advocate judging such things ex ante. I think that makes sense because the spirit of this question is "was it a good idea to elect Trump?".
It's like if we were having a big debate about whether to wear a seatbelt. The YOLOers can't be like "see, we didn't crash, told you so".
@dreev Setting aside the AI aspects—if Trump does <crazy thing> and Congress comes together to stop him and passes <big plan with lots of positive externalities that everyone agrees was the most impactful government reform ever>—is that a YES?
Re: seatbelts: the driver announces we’re taking a sudden detour down the hill, the rest of the car pulls the handbrake and decides that maybe the driver shouldn’t be able to do that and transforms into a parliamentary democracy, everyone claps… good impact?
Or simply:
Trump: “I’m going to kill us all!”
Congress: takes away the power of the president to kill us all
Trump: “See, the president shouldn’t have the power to kill us all, after all. You’re welcome, everybody!”
(Are positive outcomes from the rest of government acting against Trump positive outcomes of Trump’s second term?)
how does this resolve if top rationalists are like "yeah he did a relatively great job on almost everything, but he accelerated AI so on balance he was terrible"?
you'll probably have Robin Hanson and Roko Mijic on Trump's side
but we'll be getting toward the end of Trump's term and he'll liquidate the sovereign wealth fund to build a $5 trillion computer and rationalists generally won't be happy
@jim Yeah, it's a good question. In retrospect, maybe I should have made this exclude any impact from AI. But I'm not sure if it's fair to do that now...
@jim
1. I doubt this will happen, considering current events. If anything is going to save Trump it's gonna be some AI development that turns rationalists in his favor.
2. Roko is a clown who no one takes seriously, he's only famous because of the Basilisk and shouldn't even count as a prominent rationalist for the purposes of this market. Considering Robin Hanson has criticized rat/TPOT in the past, I don't know if he's one either.
Liron is kind of a top rationalist: https://x.com/liron/status/1886922275569721526?t=pUUaA3-OLa2nL5stPT5C9A&s=19
@JamesGrugett never heard of them
Can you please provide a few names of people you consider "prominent rationalists"?
@JamesGrugett also, I've just reread the description and I think it could be interpreted as saying that you only require 2 "prominent rationalists" to think trump was he most positively impactful in other to resolve YES.
Can you please clarify if you're looking for some kind of majority, or if a small number of heterodox voices will be sufficient?
@Fion A majority have to believe this!
This is a vibes market I don't want to get caught in too many preemptive details, or it won't judge the vibes accurately.
Like if among the top three it is mildly one way, but then including the next 10 flips it hard the other way, that seems more towards the second view. On the other hand if the top few are hardcore one way, but the next are mildly against, that seems like the former's view wins.
Also I haven't exactly chosen who it would be. We can argue that. But that is another thing that is vibes based and follows the same pattern in the previous paragraph, where there are gradations of "top rationalist".
@JamesGrugett how tf is a failed startup bro and internet yapper (+zionist zealot) a "top rationalist"? because he tweets a lot?