Intro

Yesterday, on January 31st, was the cutoff for the 2025 ACX forecasting tournament. This is the 2nd ACX contest I have entered (Metaculus username: jshapiro), but this year I’m comfortable enough with my blog to post my predictions in advance, together with a rough sketch of my model. Should be interesting and hopefully it will help me improve next year.

I did all the questions with two friends this year (the randomator and one who hasn’t chosen his true elvish name but has also forbidden me from making up my own pseudonym for him, so I will go with He Who Must Not Be Named). The biggest challenge was volume. We said we would spend 7 minutes on individual research, 5 on discussion and 3 on writing up notes. In practice we overran often and spent about 15 hours on it over the course of three weekend days (so around 24 minutes per question). Plus I spent around 3 hours on the 31st updating everything that had happened in the two weeks from our forecasting marathon (most notably the Trump inauguration & executive orders, Gaza ceasefire negotiations, the DeepSeek market panic, Russian deep sea sabotage, the declassified CIA report on the COVID origins, the Maersk advisory on shipping resumption … look, a lot is happening, ok?).

I want to do a writeup of the 2024 contest and the 2025 entry, with some data analysis and reflection, but it’s quite far down the backlog of posts. So for now I’ll just post my mostly unedited notes from the forecasting marathon for 2025 and note, regarding the 2024 contest, that Metaculus have been great regarding the shared account issue and that their summary of the situation is consistent in all the major points with my and you-know-who’s perspective (it was only the two of us doing the contest last time.

Doing this in 2025 was really fun again, thanks all around to You-Know-Who, the randomator, and to Scott/Metaculus for organizing.

Finally a disclaimer. Some questions concern market events or macroeconomic indicators. The predictions here are my own, often not held confidently. They don’t represent the views of my current employer. They are definitely not financial advice. To expand this a bit, many of these predictions are not particularly well thought-out given the time constraints. Statistically, some of what I write will be obvious nonsense. If you think that, I’d love to hear about it (it’s more impressive before the relevant question resolves ;).

Enough of this, let’s dive in.

Questions

Prediction before discussion: 10%

Prediction after discussion: 47%

Justification: Mostly independent between Polymarket and manifold, with small penalty bc CFTC hates Polymarket.

There’s a precedent for Kalshi, & some political will bc Polymarket helped the Trump campaign.

So up from 10% (by default, nothing happens) to 30% each, so 1 - (1 - 30%)² = 0.51. Pushed down a bit because the CFTC hates Polymarket.

2. Will OpenAI top the Chatbot Arena Leaderboard on December 31, 2025? 48%

Prediction before discussion: 45%

Prediction after discussion: 48%

Justification: We’ve got ChatGPT, Gemini, DeepSeek and Grok as serious contenders. Let’s give 5% and 3% respect. to Grok and DeepSeek (not the 1st mover advantage but good data, institutional barriers, Asian bias). Also 5% that Anthropic tries to enter and 1/3 that they succeed given that they try = 1.6%.

Leaving ~45% between ChatGPT and Gemini. Their ELOs are suuper close, pushing me towards coin toss.

OpenAI is a trailblazer in actual intelligence (vs commercial product). But also: politically unstable and people are leaving.

Update 2025-01-31: Giving 10% to DeepSeek because it’s more sexy than any US corporate brainwashed one. Puts new probability at 41.7%. (Update 2025-02-02: but I apparently forgot to click that into Metaculus, so 48% it is :( ).

3. Will Google’s search market share drop below 85% in 2025? 5%

Prediction before discussion: 5%

Prediction after discussion: 4.5%

Justification: Current trend indicates decrease of 1.89% per year, so they would arrive at 85% in 2.5 years (from current 89.73%).

For this to be more, we’d need someone to be really good at search. There is a market there, but not the biggest market (that’s automating labour, CRMs, emails etc.). Search is already pretty good, ads are a moat. Alternative business models, e.g. subscription, won’t scale. VC money w/o making a profit could scale, so this argument is moot.

Competitor grabs 5% market share: 4% (1/4 success conditional on 15% chance of trying again, -2% because 200M queries a day is a lot.

New startup grabs 5% market share: TikTok got 500 MAUs1 in the 1st two years. That sort of growth would not to get to 400M queries a day. Startup like this existing with the best people: 20%, success 5% -> 1%

Update 2025-01-31: Grok success could be a factor here, but it’s a wash since anyone could use their tech (including Google). I think its main competitive edge is that it’s not beholden to US corporate norms, otherwise no moat.

4. Will inflation-adjusted gas prices rise in the US in 2025? 44%

Prediction before discussion: 42%

Prediction after discussion: 44%

Justification:

It seems mean reversing. It’s close to baseline, putting me at 60% up. trend goes downwards still, so more like 50% up. But national conflict affecting oil is like 12%, so 58% up?

5. Will the 12-month percentage change in the US Consumer Price Index be lower in November 2025 than it was in November 2024? 65%

Prediction before discussion: 45%

Prediction after discussion: 65%

Justification:

Lower since Jan 2021: 12% Lower for the last 20 years: 61% -> (adjusting for COVID 20%) Lower for the whole history: 37%

Average with weights 3,2,1: 18.8%.

Adjusting upwards bc downwards trend.

60% that Musk will succeed at cutting enough to go more down. If he doesn’t, it still happens with 25% prob.

New prob = 1 - 75% * 40%.

6. Will Nvidia outperform the S&P 500 in 2025? 88.9%

Prediction before discussion:

Prediction after discussion: 93.26%

Justification:

I just ran a simple MC model and interpolated a bit between a 3 year history (AI regime) and a 20 year history (broader econ regime).

Update 2025-01-31: Due to Grok, I’m putting 15% weight on the last 20 year history number instead of 10%. Units of AI getting cheaper per unit of compute won’t make us use fewer units of compute unless we’re hitting marginal returns. My model is that we very much aren’t at marginal returns in 2025. I’m also updating with the newest market values. New number 88.9%.

7. Will Maersk resume shipping in the Red Sea in 2025? 18.3%

Prediction before discussion: 6%

Prediction after discussion: 7%

Justification:

According to Lloyd’s: Ceasefire -> peace + De-escalatory sign from Houthis: 1/4 * 1/3 = 1/12

Dock off a few points because of lead times in insurance with Lloyd’s

Update 2025-01-31: There’s a new advisory from Maersk. Also, based on the success of the Israel-Hamas ceasefire negotiations, I’m putting my chance that hostilities won’t resume by the end of 2025 up from 1/3 to 55%. No more points docking for Lloyd’s: if this model works out, we’ll resume shipping way before EOY. New p=18.3%.

8. Will the poverty rate in Argentina be lower in the first half of 2025 compared to the second half of 2023? 85%

Prediction before discussion: 93%

Prediction after discussion: 96%

Justification: TC says there’s a lot of catch up growth possible.

It’s at 52.9%, it needs to go down 11%.

It seems to be at 36.8% now.

Political system seems pretty stable.

Update 2025-01-31: No new info really, but looking at old info. The 36.8% article seems to be based on this study. While I don’t fully understand it, it might be using lagged indicators. While Milei is very competent at macro, benefits might not translate to poverty rate immediately if that’s not an explicit short-term goal. That study sure looks science though, it has confidence intervals and everything. The trend looks sus though. Also catch up growth won’t immediately translate to unemployment.Going down to 85%.

9. Will there be at least 1,000 deaths due to direct conflict between Israel and Iran in 2025? 18%

Prediction before discussion: 25%

Prediction after discussion: 24%

Justification:

Since 1979:

  • Iran Iraq War, so 1/45
  • There are active plans
  • Ongoing cold conflict
  • Trump tends yes?
  • 1/5 that something major happens that calms Israel down
  • 2/3 that they’ll do their thing if it doesn’t
  • -> 22%

Update 2025-01-31: I don’t understand my model anymore but I sure hope it’s good. Saber-rattling, no news, going down 1% for time decay for two weeks. Going down 5% points for proxy-waring.

10. Will at least twice as many deportations by U.S. ICE occur in Fiscal Year 2025 compared with Fiscal Year 2024? 58%

Prediction before discussion: 30%

Prediction after discussion: 58%

Justification: Eyeballing the growth rates, it would take the slightly more than the highest growth rate (97%) to double removals. Call that 1/5. Also, a doubling would bring deportations significantly above the all time (Obama) high.

The growth so far is likely a ramp up after covid so I would expect it to decelerate. Also the growth that is priced in is enough to be a political success, so there’s no motivation to try and double. -> 15%.

But there is political will for “Tom Homan’s mass deportation plan” to deport 20M (!) people. The guy is familiar with the process (former immigration worker, so has context). But no specific achievable goals afaict. -> 40%

Update 2025-01-31: There seems to be political move for flashy symbolic moves (pats himself on the shoulder). The Laken Riley act seems to be ending “catch-and-release”, as promised. The use of military planes is equally symbolic. But also:

U.S. Immigration and Customs Enforcement said it made an average of 710 immigration arrests daily from Thursday through Monday, up from a daily average of 311 in a 12-month period through September under President Joe Biden. If that rate holds, it would surpass ICE’s previous high mark set in the Obama administration, when daily arrests averaged 636 in 2013.

(source)

Going up to 74% based on political will and quick concrete action, even though it’s photo-copy. However, once borders shut, it will be harder to apprehend people, and that seems to constitute the majority of detentions. So maybe I should be more around 40%? Confusion. Going to 45% since I don’t have much time to dig deeper.

11. Will mifepristone become significantly restricted or illegal to prescribe for abortions across the US before 2026? 4%

Prediction before discussion: 8%

Prediction after discussion: 8%

Justification: Trump doesn’t seem to care. Supreme court says no and Trump agrees (even if it’s just a technicality). The group that advocated for it is doing lots of stuff without a lot of coordination. The drug has other uses. Doesn’t seem to go through a lot of the hoops.

Update 2025-01-31: Trump hasn’t made this a prio, going down to 4%.

12. Will the Manifold market “Did COVID-19 come from a laboratory?” be higher than 68% or lower than 28% for the majority of the final week of 2025? 22%

Prediction before discussion: 2.5%

Prediction after discussion: 8%

Justification:

Political will in China to never admit to a lab leak (anyone with evidence might be detained, intimidated or dead, if they are competent). We only get information if there was a lab leak. China seems to declassify things after about 10 years, giving a risk of at most 10% / year. It would take either declassification (4%) or a Snowden (1% given how few people are involved). Divide by 2 because we’re conditioning on lab leak. No journalists or intelligence services because evidence was destroyed. Maybe the US already knows and is keeping secret as leverage (1%) Blogger could move manifold on no evidence (5%).

Update 2025-01-31: Everything is happening in Jan, omg. CIA declassified a report (I essentially gave that 1% weight. Decicively not patting myself on the shoulder on this one).

Market isn’t updating much though? Is it just not that liquid? Also reading the comment section… idk is that market really efficient? Going up to 22% bc I’m inclined to think that the market isn’t responding and the Trump administration might find it expedient to turn this into a Story which will sway public opinion which will sway Manifold. Call me cynical.

13. Will there be a leadership change in any of United States’ foreign adversaries in 2025? 19.3%

Prediction before discussion: 16%

Prediction after discussion: 20%

Justification:

10% just from one of them dying. Add some adjustments for political risk and a 0.5% increase from probably receiving worse than UK health care.

			Age	p(death-nat)	p(death-unnat)	1-p
Xi Jimping		71	2.00%		0.75%	97.25%
Miguel Diaz-Canel	64	1.00%		0.50%	98.50%
Khameini		85	9.00%		0.75%	90.25%
Kim Jong Un		43	0.30%		1.50%	98.20%
Vladimir Putin		72	2.00%		1.50%	96.50%
Nicolas Maduro		62	1.00%		1.50%	97.50%

This mathses out to 20.12%.

Update 2025-01-31: Time decaying to 19.28%

14. Will Israel establish direct governmental control of the Gaza Strip in 2025? 4%

Prediction before discussion: 8%

Prediction after discussion: 4%

Justification: If fighting continues, 15%. If it doesn’t, 0%. Chances of it continuing: 40%. Times 70% for them wanting to do that.

15. Will a new war or a substantial escalation to a previous war kill at least 5,000 people in 2025? 48%

Prediction before discussion: 42%

Prediction after discussion: 51%

Justification:

Predicting base rate based on this data (38%). Adjust upwards because world is kinda on fire.

Update 2025-01-31: Time decaying to 48%

16. Will the US government release additional Epstein documents in 2025? 21%

Prediction before discussion: 3%

Prediction after discussion: 21%

Justification: Trump doesn’t particularly seem to care. There is no evidence that there is a list waiting for release.

It’s not been very long (no declassification) but it’s not super hot rn either.

Based on court document: Further documents exist in courts and are sealed (100%). They are sealed for unlawful reasons (50%). They will be unsealed because this judge set a good example (2/3). The unsealing won’t be contested (90%). Someone is bothered enough to do it (probably someone within a court 70%)

=> 3%

17. Will real housing prices in the US increase more in 2025 compared to 2024? 50%

Prediction before discussion: 90%

Prediction after discussion: 50%

Justification:

18. Will US residents be able to order Starbucks delivery using one of these major AI interfaces before 2026? 98%

Prediction before discussion: 96%

Prediction after discussion: 98%

Justification:

Liability concern: 70% on current model, 30% on PA model. Chance of PA model: 33% -> 30% + 11% = 44% per company. Model tells me 98%.

19. Will the Democrats be favored to win the 2028 US presidential election in the last week of 2025, according to Kalshi? 50.5%

Prediction before discussion: 50.5%

Prediction after discussion: 50.5%

Justification: CBA.

20. Will semaglutide be taken off FDA’s drug shortage list in 2025? 95%

Prediction before discussion: 85%

Prediction after discussion: 95%

Justification: In one case, there was a shortage for two years. Novo Nordisk spent $6 Bn on expanding production capacity in Nov 2023. Eli Lilly has been approved as competition (Nov 24) and they’re competent.

No regulatory barriers.

In one case, a drug stayed on the shortage list for 2 years.

Nordisk is bleeding money like hell.

21. On December 31, 2025, will Google, Meta, Amazon, Tesla, or X accept crypto as a payment? 55%

Prediction before discussion: 18%

Prediction after discussion: 55%

Justification:

6% because it’s stupid. High transaction costs and vol. The only reason to do it is to court favour with the trump administration.

So for 3 unilateralists: 9%

The remainder is Musk. 45% (ESG concerns are likely to stick around, but also there’s the crypto scam theory of the Trump Administration)

Total = 55%

22. Will any rationalist, effective altruist, or AI safety researcher go on the Joe Rogan Experience before 2026? 55%

Prediction before discussion: 56%

Prediction after discussion: 55%

Justification:

For a start, there are about 1500 twitter accounts with more than 1 M followers, according to this dodgy model.

Let’s say there are 5 rats among those. Making for a 5/1500 risk per day, for 252 bdays. (3.5 episodes a week, actually)

That gives a 56% chance, which feels dumb.

You-Know-Who gets .8 per year. Adjusted for real rats .4

23. Will an application to ban AfD be filed at the Federal Constitutional Court before 2026? 9%

Prediction before discussion: 65%

Prediction after discussion: 37%

Justification: They said they were going to do it. They will gain politically from doing it. There are changes in legislation about how plausible this is.

Update 2025-01-31: A debate took place: they don’t have the votes and put the issue into an Ausschuss w/o calling for a vote. Reading this I’m realizing I’ve been dumb. There are two motions: 20/13750 with 113 supporters of different factions, and 20/13105 with 43 supporters, all of whom are greens. The question resolves positive only if an application to ban is made based on the 2nd motion.

That’s never going to happen. I hate to go down all the way to 1% but… here I am. Going to 3% and all of that probability mass is someone in AfD doing something stupid and undeniably Nazi (even though their whole thing is very carefully not doing that, mistakes happen). Or i.g. they win (1/3) and people try to panic-ban them (1/2) after they promise to do something horrible (1/12) before they get into power (1/2)? 1%. Total 4%.

I think I should also have some weight on people trying to ban them after the election.

I dug.

Meta note: I went way deeper here than was justified given the time constraints but I find the topic interesting.

The legal case seems to be based on this open letter which is based on this document from the Institute of Human Rights.

Their whole argument is that things in the AfD is “Völkisch” (which is true and easy to prove) and that that violates GG Art. 1 of the German constitution. Now I’m not a lawyer but I think that’s just false.

Aside: Völkisch is hard to translate, and the Wikipedia page isn’t great imo. But it has this picture, which conveys the vibe. Ethno-nationalist comes close, and … I’ll shut up before I say something I regret.

My whole point is that I don't know how to caption this, sorry. It's a poster that looks very Nazi.

Back to my point which was that while the AfD often have very Nazi vibes, the IfM’s report doesn’t make the case why those vibes are incompatible with the Basic Democratic Order.

Here some translations. GG Art 1:

(1) The dignity of the individual is inviolable. To respect and protect it is the duty of all state authority. (2) The German people therefore commit to inalienable and inviolable human rights as the foundation of every human community, of peace, and of justice in the world. (3) The following fundamental rights [Art. 1 of course means that this is at the start of the constitution, so “the following” means the entire constitution] bind the legislature, the executive, and the judiciary as directly applicable law.

Excerpt from the AfD Manifesto.

Our identity is shaped by our German language, our values, our history, and our culture. The latter are closely connected to Christianity, the Enlightenment, our artistic and scientific works. Our identity determines the fundamental values that are passed down from generation to generation. The German Leitkultur (leading culture) describes our value consensus, which is identity-forming for our people and distinguishes us from others. It ensures the cohesion of society and is a prerequisite for the functioning of our state. The community-building effect of German culture is the foundation of our Basic Law and cannot be replaced by a form of constitutional patriotism. Cultural relativism and multiculturalism lead to a coexistence and opposition of parallel societies, which lack common values for living together. In such a fragmented society, conflicts arise that are hardly manageable. The AfD will not allow Germany to lose its traditional culture out of a misguided tolerance for Islam.

There’s also this thing written by 17 constitutional scholars which seems more measured but also says that there is a legal basis for a ban.

It makes two points. The first one is about Völkisch Ideology again, but quoting the constitutional court when it banned the NPD saying that that’s unconstitutional maybe BVerfGE 144, 20 [207] [NPD-II-Verbotsverfahren, 2017]? I’ve read the excerpt and I think the government can’t demand “The german people must give everything to The Cause™” but pushing a primary culture would be fine from that standpoint. Excerpt:

The subject quality of the individual is associated with a social claim to value and respect, which prohibits degrading the individual to a “mere object” of state action (cf. BVerfGE 122, 248 [271]).

Even though this “object formula” may be limited in its effectiveness (cf. BVerfGE 109, 279 [312]; critically Dreier, loc. cit., Art. 1 para. 1 note 55; Höfling, loc. cit., Art. 1 note 15 – each with references), it is suitable for identifying violations of human dignity wherever the subject quality of the individual and the resulting claim to respect are fundamentally called into question (this is also the conclusion of Dreier, loc. cit., Art. 1 para. 1 notes 60 ff.). This is especially the case with any notion of an original and thus unconditional priority of a collective over the individual. 2

But the BVerfGE points to limitations of this argument in this very formula.

Second, they say that calling the media fake news and calling the government a shadowy cabal of elite oligarchs (I’m being very loose with my translations here) is a threat to a well-functioning democracy. I kind of buy this, actually.

24. Will the US unemployment rate in November 2025 be below the rate in November 2024? 21%

Prediction before discussion: 18%

Prediction after discussion: 18%

Justification: How low are we now? Lower than 77%.

But we’re in a weird regime. Never before has there been this long a gradual growth.

Putting 1/4 weight that this is because of AI, in which case it will continue. In the other case, I’m putting 2/3 rds on it continuing, giving a total of 75%.

Actually, if split by education, it would seem that all education levels face the same trends, where AI would disproportionately affect high education. This puts AI down to 1/6th, so I’m down to 72%.

There is also the explanation that software has stopped eating the world, and we’re gradually reversing the over-hiring / momentum of that.

Two separate mechanisms bring it upwards, so I’m going with 85%.

Update 2025-01-31: Found this3. Specifically this quote:

he rise in the unemployment rate this summer raised concern about whether labour market re-balancing would go too far. But we attribute the increase mainly to short-term frictions associated with absorbing rapid labour force growth caused by the immigration surge, not to weak labour-demand —after all, job openings are high and the layoff rate is low

Big if true, but even then the effect will take some time according to GS. Updating upwards by 3% to 21%.

25. Will US federal interest rates at the end of 2025 be lower than at the end of 2024? 85%

Prediction before discussion: 92%

Prediction after discussion: 85%

Justification:

77% base rate, they said it would lower it a bit more. Market seems to agree. Fed is reliable.

SPX has about 3% bad draw-down years. And there’s 12% of bubbles. We’ll raise rates only to stave off a bubble.

26. Will major components of the Affordable Care Act be repealed or curtailed before 2026? 10%

Prediction before discussion: 17%

Prediction after discussion: 12%

Justification:

Complicated to satisfy, so far he’s stuck to symbolic moves. It’s politically a bad idea to actually do it, he’s just catering to libertarians. He’s been inconsistent on the issue, it’s just “boo Obama!”.

Update 2025-01-31: I don’t want to study the US healthcare system. He did two executive orders affecting Medicaid, none of them radical (although premiums were increased, which shows he does indeed dislike it). There is a good chance he’ll leave ACA at that, because healthcare is popular among voters actually, so going down to 10%.

27. Will Iran possess a nuclear weapon before 2026? 38%

Prediction before discussion: 68%

Prediction after discussion: 38%

Justification:

If no intervention, almost certainly. They’re doing breakout. Successful intervention from US through CIA op or outright war: 5% Successful intervention in joint op with Israel: 1/2 * 80% = 40% Mess up because tech is hard; 7%

Tech is hard, so 20% they are not ready.

28. Will there be a bilateral ceasefire in the Russo-Ukraine conflict before 2026? 20.7%

Prediction before discussion: 28%

Prediction after discussion: 28%

Justification:

I put 30% last year and that was low. Western support is wavering and the west is getting distracted.

But also, Trump is willing to trade support for resources, which would prolong the conflict.

Update 2025-01-31: Time decay to 26.7%. Also the cable stuff seems escalatory and unafraid of Trump, so going down 5% points to 21.7%.

29. Will Benjamin Netanyahu remain Prime Minister of Israel throughout 2025? 72.5%

Prediction before discussion: 78%

Prediction after discussion:

Justification:

His term is about to end, and you don’t usually resign right before unless something horrendous happens. What would that even be at this point? There is a lot of war going on, which is high variance but probably helps popularity. There could be tail events with war that oust him. 3% He seems remarkably stable. He has 2% of just dying. He’s survived a recent war related attempt to oust him. He might resign after the war with Hamas is over? 50% he does it, 30% it’s over. 2% for random politics BS.

3% + 2% + 2% + 15%

Update 2025-01-31: 55% peace and 50% he resigns given peace. New prob: 72.5%.

30. Will Argentina’s year-over-year inflation be below 30% for December 2025? 92%

Prediction before discussion: 38%

Prediction after discussion: 92%

Justification:

Based on trends, it would have to drop more than the drop from max in April to Dec. So starting point 60%.

This is optimistic, since looking at the data it seems to have mostly stabilized, so maybe solving hyperinflation is done, but improving the economy is a separate problem that could take lower.

The currency peg change will need to succeed, but if it does it will help.

31. Will Donald Trump have a positive net favorability rate on December 20, 2025, according to 538? 12%

Prediction before discussion: 12%

Prediction after discussion: 12%

Justification:

Everyone hates the incumbent towards the end. He also doesn’t actually give a damn about reelection. He’s never been popular, and I’m not sure what has changed.

The trend points upwards, but that might well be honeymoon.

32. Will the US eliminate the Department of Education through law or presidential reorganization before 2026? 1.5%

Prediction before discussion: 5%

Prediction after discussion: 1.5%

Justification:

No elections, but republicans don’t have the vote. 7 democrats missing and it looks bad to educated elites who won’t educate themselves about the implications? Also everyone is afraid of what Texas would do etc. -> 5% (This is because the Senate must agree).

Simple majority for reorganization but senate must agree, so it’s the same pathway.

Update 2025-01-31: He’s passed EOs regarding the EOD but hasn’t talked about dismantling it. This fits my model, no update.

33. Will Jerome Powell cease to hold the office of Chair of the Federal Reserve before 2026? 2%

Prediction before discussion: 5.6%

Prediction after discussion: 2%

Justification: He said he won’t fire Powell and his term ends anyways. So 0% + chance of death (1%).

Chance to substantially modify fed 5% (it was designed to be sturdy). Chance that that would dethrone Powell: 1/5 (I just don’t think that’s how it usually works). 1.6%

34. Will the US Congress pass a bill implementing Trump’s tariff policies in 2025? 2.5%

Prediction before discussion: 2.5%

Prediction after discussion: 2.5%

Justification:

This is an incredibly dumb idea with massive ramifications. The markets are long Trump so seem to not be pricing this as happening.

Trump has been great for the US economy.

Besset says that 20% broad tariffs is exaggerated and a negotiating position. => 2%

Maybe through DBCFs? Seems unlikely, not popular, industry push-back. 0.5%.

Update 2025-01-31: GS gives global sanctions 10%3. I think the inauguration speech w/ establishing an external revenue service makes DBCFs more likely, I’ll give them 2%. I’ll also bump the base rate for the stupid thing up by 1%, for a total of 5%. I still think that most of the taxation will be done on a country-by-country basis.

37. Will Elon Musk cease to be an advisor to Donald Trump and face public criticism from Donald Trump before 2026? 15%

Prediction before discussion: 28%

Prediction after discussion: 16%

Justification:

I can see ~20 people on the list of Trump survivors which I recognize. I can see ~7 people who got fired who I recognize. This is all biased as fuck but gives me a 26% chance of getting fired.

Musk is more likely to be disagreeable and my model is that really insufferable people are really insufferable. And Trump doesn’t care about outcomes the same way that capitalism does.

On the other hand: BTC theory of everything.

Also Musk could meme in a way that is good for Trump.

Trump needs twitter and Musk’s platform to some extent.

Actually it’s only 4 people I know who’ve lasted less than a year.

Update 2025-01-31: Time decay to 15%.

38. Will Elon Musk be the world’s richest person on December 31, 2025? 98.8%

Prediction before discussion: 98.8%

Prediction after discussion: 98.8%

Justification:

Give him .2% of dying.

1% for two unlikely things happening together (tesla haves in value and Space-X looses business to Jeff; he goes to prison and Trump doesn’t pardon him)

  1. Monthly active users. ↩

  2. All translations here by ChatGPT. ↩

  3. Goldman Sachs 2025 US Economic Outlook. ↩ ↩2