Hide table of contents

Summary

  • I think the prior probability of this being the most important century is less than 1 % based on i) the self-sampling assumption (SSA), ii) a loguniform distribution of the effective computation this century as a fraction of that throughout all time, and iii) all effective computation being more than 10^44 times as large as that this century.
  • The bigger the influentialness of the future, the lower the probability of this century being influential.

Methods

In agreement with section 2 of Mogensen 2022, the self-sampling assumption (SSA) says the prior probability of being the most influential person among a population of size N, comprising past, present and future people, is the expected value of 1/N (not the reciprocal of the expected value of N). To be more general, I have estimated the prior probability of this being the most important century from the expected value of effective computation this century as a fraction of that throughout all time. By effective computation, I mean one which counterfactually increases/decreases welfare, regardless of whether it comes from biological beings or artificial intelligence (AI).

Artificial general intelligence (AGI) might soon cause lock-in to a certain extent. If this does not involve total loss of value[1], the vast majority of computation could still happen in the longterm future, but one would not be able to counterfactually increase/decrease welfare.

In addition, this might be the most important century owing to high risk of human extinction. However, a priori, one should arguably expect such risk to also be proportional to the expected value of effective computation this century as a fraction of that throughout all time. Along the same lines, I wonder whether the vast majority of technological progress will happen in the longterm future.

I modelled the effective computation this century as a fraction of that throughout all time as a loguniform distribution, which is a good uninformative prior. Its expected value is (b - a)/(ln(b) - ln(a)), where a and b = 1 are its minimum and maximum. As a approaches 0 for b = 1, the expected value tends to 1/ln(1/a). In other words, the prior probability of this being the most important century is inversely proportional to the logarithm of the maximum effective computation throughout all time as a fraction of that this century. The bigger the influentialness of the future, the lower the probability of this century being influential.

Results

The results are below. My calculations are in this Sheet.

Discussion

The future can hold at least as many as 10^54 lives according to Newberry 2021 (Table 1), and this century can have less than 10 billion (10^10) lives. Consequently, 1/a is lower than 10^(54 - 10) = 10^44, which suggests a prior probability of this being the most important century of less than 0.987 %.

Note I would have (wrongly) arrived at a much lower probability based on the reciprocal of the expected effective computation throughout all time as a fraction of that this century. For the conservative estimate of 10^28 expected future lives given in Newberry 2021 (Table 3), which is at least 10^(28 - 10) = 10^18 times as many as the lives this century, I would get a prior probability of 10^-18. Much lower than 1 %! 10^-18 is the kind of value which would result from following the formulation arguably implied in MacAskill 2020[2], but it is not in agreement with the SSA, as explained in section 2 of Mogensen 2022

Acknowledgements

Thanks to David Denkenberger for feedback on the draft.

  1. ^

     I do not think AGI causing human extinction would necessarily lead to a permanent loss of value.

  2. ^

     William MacAskill provides numbers with respect to time. Consequently, since population size increases throughout time, all lives as a fraction of those this century is much smaller than all valuable time as a fraction of 100 years. So William gets a probability higher than 10^-18 for this being the most important century. Nevertheless, William agrees population is also relevant:

    However, for the purposes of true action-relevance, the influentialness of a time is not quite what we’re looking for. Even if we assume that now is the most influential time, because there are available opportunities to safeguard the longrun future, a rural farmer in Central African Republic would simply not be able to access those opportunities, and so for that person the question of how influential the present time is is neither here nor there. So we can generalize Trammell’s model slightly by talking about person-times rather than times: rather than being indexed to a particular time, each term in his model should be indexed to a particular person-time [i.e. life].

8

0
0

Reactions

0
0

More posts like this

Comments2


Sorted by Click to highlight new comments since:

Artificial general intelligence (AGI) might soon cause lock-in to a certain extent. If this does not involve total loss of value[1], the vast majority of computation could still happen in the longterm future, but one would not be able to counterfactually increase/decrease welfare.

Sure, you don't buy this but shouldn't you still account for it in your prior? or is this just "a prior if we discount extinction"?

Hi Nathan,

The prior is supposed to account for extinction risk.

However, a priori, one should arguably expect such risk to also be proportional to the expected value of effective computation this century as a fraction of that throughout all time [?]. Along the same lines, I wonder whether the vast majority of technological progress will happen in the longterm future.

Curated and popular this week
LintzA
 ·  · 15m read
 · 
Cross-posted to Lesswrong Introduction Several developments over the past few months should cause you to re-evaluate what you are doing. These include: 1. Updates toward short timelines 2. The Trump presidency 3. The o1 (inference-time compute scaling) paradigm 4. Deepseek 5. Stargate/AI datacenter spending 6. Increased internal deployment 7. Absence of AI x-risk/safety considerations in mainstream AI discourse Taken together, these are enough to render many existing AI governance strategies obsolete (and probably some technical safety strategies too). There's a good chance we're entering crunch time and that should absolutely affect your theory of change and what you plan to work on. In this piece I try to give a quick summary of these developments and think through the broader implications these have for AI safety. At the end of the piece I give some quick initial thoughts on how these developments affect what safety-concerned folks should be prioritizing. These are early days and I expect many of my takes will shift, look forward to discussing in the comments!  Implications of recent developments Updates toward short timelines There’s general agreement that timelines are likely to be far shorter than most expected. Both Sam Altman and Dario Amodei have recently said they expect AGI within the next 3 years. Anecdotally, nearly everyone I know or have heard of who was expecting longer timelines has updated significantly toward short timelines (<5 years). E.g. Ajeya’s median estimate is that 99% of fully-remote jobs will be automatable in roughly 6-8 years, 5+ years earlier than her 2023 estimate. On a quick look, prediction markets seem to have shifted to short timelines (e.g. Metaculus[1] & Manifold appear to have roughly 2030 median timelines to AGI, though haven’t moved dramatically in recent months). We’ve consistently seen performance on benchmarks far exceed what most predicted. Most recently, Epoch was surprised to see OpenAI’s o3 model achi
Dr Kassim
 ·  · 4m read
 · 
Hey everyone, I’ve been going through the EA Introductory Program, and I have to admit some of these ideas make sense, but others leave me with more questions than answers. I’m trying to wrap my head around certain core EA principles, and the more I think about them, the more I wonder: Am I misunderstanding, or are there blind spots in EA’s approach? I’d really love to hear what others think. Maybe you can help me clarify some of my doubts. Or maybe you share the same reservations? Let’s talk. Cause Prioritization. Does It Ignore Political and Social Reality? EA focuses on doing the most good per dollar, which makes sense in theory. But does it hold up when you apply it to real world contexts especially in countries like Uganda? Take malaria prevention. It’s a top EA cause because it’s highly cost effective $5,000 can save a life through bed nets (GiveWell, 2023). But what happens when government corruption or instability disrupts these programs? The Global Fund scandal in Uganda saw $1.6 million in malaria aid mismanaged (Global Fund Audit Report, 2016). If money isn’t reaching the people it’s meant to help, is it really the best use of resources? And what about leadership changes? Policies shift unpredictably here. A national animal welfare initiative I supported lost momentum when political priorities changed. How does EA factor in these uncertainties when prioritizing causes? It feels like EA assumes a stable world where money always achieves the intended impact. But what if that’s not the world we live in? Long termism. A Luxury When the Present Is in Crisis? I get why long termists argue that future people matter. But should we really prioritize them over people suffering today? Long termism tells us that existential risks like AI could wipe out trillions of future lives. But in Uganda, we’re losing lives now—1,500+ die from rabies annually (WHO, 2021), and 41% of children suffer from stunting due to malnutrition (UNICEF, 2022). These are preventable d
Rory Fenton
 ·  · 6m read
 · 
Cross-posted from my blog. Contrary to my carefully crafted brand as a weak nerd, I go to a local CrossFit gym a few times a week. Every year, the gym raises funds for a scholarship for teens from lower-income families to attend their summer camp program. I don’t know how many Crossfit-interested low-income teens there are in my small town, but I’ll guess there are perhaps 2 of them who would benefit from the scholarship. After all, CrossFit is pretty niche, and the town is small. Helping youngsters get swole in the Pacific Northwest is not exactly as cost-effective as preventing malaria in Malawi. But I notice I feel drawn to supporting the scholarship anyway. Every time it pops in my head I think, “My money could fully solve this problem”. The camp only costs a few hundred dollars per kid and if there are just 2 kids who need support, I could give $500 and there would no longer be teenagers in my town who want to go to a CrossFit summer camp but can’t. Thanks to me, the hero, this problem would be entirely solved. 100%. That is not how most nonprofit work feels to me. You are only ever making small dents in important problems I want to work on big problems. Global poverty. Malaria. Everyone not suddenly dying. But if I’m honest, what I really want is to solve those problems. Me, personally, solve them. This is a continued source of frustration and sadness because I absolutely cannot solve those problems. Consider what else my $500 CrossFit scholarship might do: * I want to save lives, and USAID suddenly stops giving $7 billion a year to PEPFAR. So I give $500 to the Rapid Response Fund. My donation solves 0.000001% of the problem and I feel like I have failed. * I want to solve climate change, and getting to net zero will require stopping or removing emissions of 1,500 billion tons of carbon dioxide. I give $500 to a policy nonprofit that reduces emissions, in expectation, by 50 tons. My donation solves 0.000000003% of the problem and I feel like I have f