TLDR: It seems useful for someone who cares about the long-term future to compare themselves and the current time period to future people and times.
Five days ago, I told my parents I was going to look into whether this is currently the most important time in history. My parents told me not to look into silly questions and to do something with my life instead.
In retrospect, I’m happy with my decision-making process. However, I could’ve explained myself better.
The meaning of hinge of history and why it’s relevant
To start, I should’ve explained that I was looking into Will MacAskill’s article discussing whether it’s currently the “hinge of history.” (I’ll also discuss material from his original post on the hinge of history.) MacAskill states that would mean, “we [present humans] are among the very most influential people ever, out of a truly astronomical number of people who will ever live.”[1]
So what does it mean to be influential? MacAskill wrote, “I will focus on…how much expected good one can do with the direct expenditure (rather than investment) of a unit of resources at a given time. I will call this the influentialness of a time.”
So to be one of the most influential people, one needs to be able to do enough expected good through direct expenditures.
Good
MacAskill doesn’t explicitly define good. I don’t know if he thinks there is an objective good.
As I’ve said, I think morality is subjective. So I believe what’s good would be different for everyone. To me, doing the most good would be maximizing utility (probably happiness) for as many beings as possible while minimizing utility inequality.
I’m skeptical MacAskill would agree with my definition. He helped create an online textbook about utilitarianism, so he may not care about inequality.
Since I think what’s good is subjective, that means I think there’s no objective hinge of history. Whenever I write about whether I think it’s the hinge of history, I’m referring to the hinge of history to me. (I use the terms hinge of history and most influential time interchangeably.)
Expected good
The term expected good clarifies that people can’t 100% know the consequences of their actions. They can only make decisions based on the information available to them.
MacAskill acknowledges important events could be complete accidents. All of our actions have ripple effects. For example, pretend someone facetiously told five-year-old Adolf Hitler that he couldn’t listen to anyone, so he should become a dictator. Imagine Hitler didn’t recognize the comment was a joke and was instead inspired to become a dictator. While that comment would’ve been incredibly important, I’d guess its expected value would be slightly negative.
I’d care about the odds an unlikely event occurs if that event would be extremely bad or good. But I’d guess there’s essentially zero chance that a joke causes a five-year-old to become a fascist dictator. And the joke could make a much smaller negative impact because it hurts the five-year-old’s feelings. If the speaker thought their joke would entertain other people in the room, it arguably has a positive expected utility.
Direct Expenditure
MacAskill contrasts direct expenditure, which he also terms consumption, with investment. He defines an investment as an activity intended to impact the future indefinitely. This could involve investing money or spreading a movement’s ideology.
The term direct expenditure seems to be used to refer to any activity that’s not intended to abstractly benefit the long term. Even working on AI Safety research to prevent AI from causing extinction over the next few hundred years would be a direct expenditure.
MacAskill states people should generally invest to maximize the resources available to prepare for the hinge of history. So when it’s the hinge of history, people would have more resources to consume.
In reality, things are more complicated. I doubt I’ll ever be certain that it’s the hinge of history. If one thinks there’s a 40% chance it’s the hinge of history, they could spend 40% of their resources on direct expenditures. Of course, it would be tough for anyone who has one job to spend 40% of their resources on an activity. Someone could coordinate with a community, such as the effective altruism community instead. The community could encourage 40% of its members to work on direct expenditures.
I think it’s worth thinking about what everyone else in the world is doing too. For example, someone could believe there’s a .01% chance it’s the hinge of history because of AI. But, since way less than .01% of the population is working on AI Safety, it would feel worthwhile for them to work on it.
Lastly, I don’t think someone who invests rather than consumes is less impactful. I doubt MacAskill feels this way either. His point seems to be that investing generally has a less direct impact than a direct expenditure. And he chooses to use the word influential to refer to direct impact.
How Long Is The Hinge Of History
There is no set amount of time when the hinge of history takes place. In MacAskill’s original post, he asks what century represents the hinge of history. But there would be a most influential decade or minute too.
It seems most useful to ask questions such as if I expect to live about 50-70 more years, will the next 60 years be the most influential 60 years. If I don’t think so, I’d be more likely to try to earn money or work on a social or political movement in my career. I’d donate the money I earn to a fund that promises to make optimal decisions over the long term. For example, the Founders Pledge Patient Philanthropy Fund. I’d also consider donating to meta-causes such as building effective altruism through the Effective Altruism Infrastructure Fund. Granted, I’d have to factor in the risk that I’d still trust the non-profit to be competent over the long-term future.
If I did think the next 60 years would be the most influential 60 years, I’d ask myself about shorter time-frames. If I thought the hinge of history would occur within 10 years, I’d focus on maximizing my impact during that time. Let’s pretend I thought an artificial superintelligence would be developed by 2032 and that would represent the hinge of history. In that case, I’d consider working as a software engineer for an AI Safety organization, being a personal assistant for an AI Safety researcher or earning and donating as much money as possible to AI safety organizations. If I thought the most influential time would be between 10-60 years from now, I’d focus on direct expenditures that pay off during that time.
Hinge of the Future
In MacAskill’s original post, he acknowledges the hinge of history could’ve been in the past. But I don’t know how to change the past. I’m focused on making the future go as well as possible. It only seems worthwhile to study the past to help predict the future. (I suspect MacAskill agrees with this point and just didn’t explicitly state it.)
The Relevant Question
MacAskill’s original post asks, “Are we living in the most influential time ever?” MacAskill’s updated post asks, “Are we the most influential people ever?”
Just knowing the answer to both questions is useless trivia. But I think it’s worth trying to answer both of those questions to help answer the more relevant question of, “What should we do with our lives?”
The lesson I took from reading about the hinge of history is that I didn’t think enough about comparing the present to future time periods and myself to future people.
I’d thought about how my strengths and weaknesses compare to people alive today, but never in a historical context. MacAskill points out that a well-educated European in the 1600s believed mice grew out of straw.[2] Maybe I’ll seem equally dumb compared to people in a few hundred years?
MacAskill illustrates how the same logic applies to determining the most important causes. The consensus that climate change is an issue didn’t emerge until the 1970s. The idea of nuclear winter was developed in the 1980s. And the importance of deep learning for AI safety research wasn’t recognized until the 2010s. Maybe we’re not yet aware of the most important problem in the future?
While I think morality is subjective, I’ve acknowledged that if I was born in the 1600s I could’ve been racist, sexist and intolerant of other religions. Moral views are still changing. Sixty percent of the population opposed gay marriage in 2004. So I’m open to investing to help future people because they’d probably convince me to change my morals.
But maybe I shouldn’t? Maybe it’s the hinge of history now? Gwern commented on MacAskill’s post that people have underestimated the importance of their time period before. Isaac Newton thought he rediscovered knowledge prior civilizations had lost because comets destroyed them. Lucretius didn’t think the technological advances of the Greeks and Romans were especially notable because he incorrectly assumed humans hadn’t already lived for hundreds of thousands of years.
I’ll start trying to quantify the odds it’s currently the hinge of history in my next post.
I suspect MacAskill meant to write “could ever live” instead of “will ever live. He acknowledges that he thinks there’s a .01 to 1% chance of human extinction this century.
Grappling With The Hinge Of History Part 1: What It Means And Why I Care
TLDR: It seems useful for someone who cares about the long-term future to compare themselves and the current time period to future people and times.
Five days ago, I told my parents I was going to look into whether this is currently the most important time in history. My parents told me not to look into silly questions and to do something with my life instead.
In retrospect, I’m happy with my decision-making process. However, I could’ve explained myself better.
The meaning of hinge of history and why it’s relevant
To start, I should’ve explained that I was looking into Will MacAskill’s article discussing whether it’s currently the “hinge of history.” (I’ll also discuss material from his original post on the hinge of history.) MacAskill states that would mean, “we [present humans] are among the very most influential people ever, out of a truly astronomical number of people who will ever live.”[1]
So what does it mean to be influential? MacAskill wrote, “I will focus on…how much expected good one can do with the direct expenditure (rather than investment) of a unit of resources at a given time. I will call this the influentialness of a time.”
So to be one of the most influential people, one needs to be able to do enough expected good through direct expenditures.
Good
MacAskill doesn’t explicitly define good. I don’t know if he thinks there is an objective good.
As I’ve said, I think morality is subjective. So I believe what’s good would be different for everyone. To me, doing the most good would be maximizing utility (probably happiness) for as many beings as possible while minimizing utility inequality.
I’m skeptical MacAskill would agree with my definition. He helped create an online textbook about utilitarianism, so he may not care about inequality.
Since I think what’s good is subjective, that means I think there’s no objective hinge of history. Whenever I write about whether I think it’s the hinge of history, I’m referring to the hinge of history to me. (I use the terms hinge of history and most influential time interchangeably.)
Expected good
The term expected good clarifies that people can’t 100% know the consequences of their actions. They can only make decisions based on the information available to them.
MacAskill acknowledges important events could be complete accidents. All of our actions have ripple effects. For example, pretend someone facetiously told five-year-old Adolf Hitler that he couldn’t listen to anyone, so he should become a dictator. Imagine Hitler didn’t recognize the comment was a joke and was instead inspired to become a dictator. While that comment would’ve been incredibly important, I’d guess its expected value would be slightly negative.
I’d care about the odds an unlikely event occurs if that event would be extremely bad or good. But I’d guess there’s essentially zero chance that a joke causes a five-year-old to become a fascist dictator. And the joke could make a much smaller negative impact because it hurts the five-year-old’s feelings. If the speaker thought their joke would entertain other people in the room, it arguably has a positive expected utility.
Direct Expenditure
MacAskill contrasts direct expenditure, which he also terms consumption, with investment. He defines an investment as an activity intended to impact the future indefinitely. This could involve investing money or spreading a movement’s ideology.
The term direct expenditure seems to be used to refer to any activity that’s not intended to abstractly benefit the long term. Even working on AI Safety research to prevent AI from causing extinction over the next few hundred years would be a direct expenditure.
MacAskill states people should generally invest to maximize the resources available to prepare for the hinge of history. So when it’s the hinge of history, people would have more resources to consume.
In reality, things are more complicated. I doubt I’ll ever be certain that it’s the hinge of history. If one thinks there’s a 40% chance it’s the hinge of history, they could spend 40% of their resources on direct expenditures. Of course, it would be tough for anyone who has one job to spend 40% of their resources on an activity. Someone could coordinate with a community, such as the effective altruism community instead. The community could encourage 40% of its members to work on direct expenditures.
I think it’s worth thinking about what everyone else in the world is doing too. For example, someone could believe there’s a .01% chance it’s the hinge of history because of AI. But, since way less than .01% of the population is working on AI Safety, it would feel worthwhile for them to work on it.
Lastly, I don’t think someone who invests rather than consumes is less impactful. I doubt MacAskill feels this way either. His point seems to be that investing generally has a less direct impact than a direct expenditure. And he chooses to use the word influential to refer to direct impact.
How Long Is The Hinge Of History
There is no set amount of time when the hinge of history takes place. In MacAskill’s original post, he asks what century represents the hinge of history. But there would be a most influential decade or minute too.
It seems most useful to ask questions such as if I expect to live about 50-70 more years, will the next 60 years be the most influential 60 years. If I don’t think so, I’d be more likely to try to earn money or work on a social or political movement in my career. I’d donate the money I earn to a fund that promises to make optimal decisions over the long term. For example, the Founders Pledge Patient Philanthropy Fund. I’d also consider donating to meta-causes such as building effective altruism through the Effective Altruism Infrastructure Fund. Granted, I’d have to factor in the risk that I’d still trust the non-profit to be competent over the long-term future.
If I did think the next 60 years would be the most influential 60 years, I’d ask myself about shorter time-frames. If I thought the hinge of history would occur within 10 years, I’d focus on maximizing my impact during that time. Let’s pretend I thought an artificial superintelligence would be developed by 2032 and that would represent the hinge of history. In that case, I’d consider working as a software engineer for an AI Safety organization, being a personal assistant for an AI Safety researcher or earning and donating as much money as possible to AI safety organizations. If I thought the most influential time would be between 10-60 years from now, I’d focus on direct expenditures that pay off during that time.
Hinge of the Future
In MacAskill’s original post, he acknowledges the hinge of history could’ve been in the past. But I don’t know how to change the past. I’m focused on making the future go as well as possible. It only seems worthwhile to study the past to help predict the future. (I suspect MacAskill agrees with this point and just didn’t explicitly state it.)
The Relevant Question
MacAskill’s original post asks, “Are we living in the most influential time ever?” MacAskill’s updated post asks, “Are we the most influential people ever?”
Just knowing the answer to both questions is useless trivia. But I think it’s worth trying to answer both of those questions to help answer the more relevant question of, “What should we do with our lives?”
The lesson I took from reading about the hinge of history is that I didn’t think enough about comparing the present to future time periods and myself to future people.
I’d thought about how my strengths and weaknesses compare to people alive today, but never in a historical context. MacAskill points out that a well-educated European in the 1600s believed mice grew out of straw.[2] Maybe I’ll seem equally dumb compared to people in a few hundred years?
MacAskill illustrates how the same logic applies to determining the most important causes. The consensus that climate change is an issue didn’t emerge until the 1970s. The idea of nuclear winter was developed in the 1980s. And the importance of deep learning for AI safety research wasn’t recognized until the 2010s. Maybe we’re not yet aware of the most important problem in the future?
While I think morality is subjective, I’ve acknowledged that if I was born in the 1600s I could’ve been racist, sexist and intolerant of other religions. Moral views are still changing. Sixty percent of the population opposed gay marriage in 2004. So I’m open to investing to help future people because they’d probably convince me to change my morals.
But maybe I shouldn’t? Maybe it’s the hinge of history now? Gwern commented on MacAskill’s post that people have underestimated the importance of their time period before. Isaac Newton thought he rediscovered knowledge prior civilizations had lost because comets destroyed them. Lucretius didn’t think the technological advances of the Greeks and Romans were especially notable because he incorrectly assumed humans hadn’t already lived for hundreds of thousands of years.
I’ll start trying to quantify the odds it’s currently the hinge of history in my next post.
I suspect MacAskill meant to write “could ever live” instead of “will ever live. He acknowledges that he thinks there’s a .01 to 1% chance of human extinction this century.
He cites page 6 of this book. I also found this article on the topic.