Asymmetric Evil

     One of the constant themes in my Losing Money Effectively writing has been the idea of asymmetry. The asymmetric payoff from playing defense is a common one: where if you prevent a nuclear war probably no one knows and you don't get billions of dollars, but if you develop AGI or some other groundbreaking (and dangerous) technology you become one of the richest and most powerful people in history. Sort of in a similar theme, I was recently thinking about the potential that people have to cause great societal harm. If you live your life to the fullest, you may be able to provide enough utility to others to equate to a dozen lives. Maybe you have children, treat them really well, be nice to your coworkers, be a great husband/wife, and donate quite a bit of money to charity. Barring some exceptional luck, you probably won't be a billionaire or famous, and thus your sphere of influence is likely to remain small. If you aren't born into a wealthy family, even with a perfect work ethic you are unlikely to reach a high enough level of status to cause large levels of change.

    Unfortunately, being a good person doesn't change society, except for at the margins. Being a bad person, in contrast, can have a really, really negative impact. If you engineer a pandemic, or shoot up a public area, or assassinate the right person, you can cause quite a bit of harm over a large sphere of influence. Maybe you shoot JFK, but if you want to cause real long-term human suffering for thousands or even millions, shoot Lincoln or Archduke Franz Ferdinand. A motivated terrorist can kill quite a bit of people, and a small group proved in 2001 that with simple planning you can kill thousands of people and spark a war that kills hundreds of thousands of people. Nineteen terrorists, a hundred thousands deaths. There's not many nineteen person nonprofits that save hundreds of thousands of lives on their own.

    This is, of course, a massive problem. In a lot of ways, human society is built on trust. Trust that the overwhelming majority (99.999999% of people) are not evil, or at least not both smart and evil. The data seems to back this up for the most part, as I don't live in constant fear whenever I go to a concert or a public park. Sure, the fear may be there, but for the most part it is irrational. Still, I think this concept of asymmetric evil is a very understaffed problem. To prevent mass shootings there are advocates for gun control (which I support), but for increased anti-terrorism efforts we often see a decrease in human freedom. It's hard to be a serial killer in an authoritarian regime that tracks your every move, but that does not mean I'd trade aggregate human freedom for a world with a handful of less serial killers. Also, we saw with the Patriot Act that a lot of times these "safety" measures actually do more harm then good. 

    This is an important concern of mine, and I do think we could do a few things better. First, we in no way shape or form should trust sensitive technological information that could lead to mass deaths to the public domain. If the government finds out how to create a super virus, they should not open source that information. This seems obvious, but for some reason (looking at you Australian smallpox researchers), it has to be said. Next, we shouldn't trust any individual or small group of individuals with massive amounts of power. Any weapons of mass destruction plans should have the required redundancy attached, less we find ourselves in the world of Dr. Strangelove. Third, we should be very cognizant of how fragile society is. There are probably reasonably easy ways to trigger a societal collapse (financial meltdown, kill the right world leader and blame a different world power), so we should be extremely diligent when building institutions and planning for the worst case scenario. In the meantime, we should acknowledge that our "good person" impact will likely be small and continue to stay the course anyway.

Sacrifice vs. Uncertainty

     For some reason, in my fiction writing at least, I can't stop writing about the idea of sacrifice. Maybe it is my Irish Catholic upbringing, or maybe it was the fact that I've read a lot of Cioran, but regardless the idea intrigues me endlessly. Here is my question:

    Would you take a bullet for democracy in the United States? You get shot in the chest, you may or may not live.

    If you don't take the bullet, the US becomes a dictatorship of North Korean proportions. If you jump in front of the bullet, nothing changes in the US. There are plenty of interesting derivations to this question. Maybe you throw in that democracy falls in 100 years, so you and your family will be fine. Or you change the prompt to reflect some other cause. Well, many American soldiers have lost their lives fighting for certain American ideals, democracy being the foremost. Probably the coolest part of the US is our history of standing up against tyranny, and taking a look at the world stage shows us as mostly alone. It's pretty crazy to me that the American experiment actually worked, and I don't see any obvious civilizational collapse on the horizon despite what the media says. I'm taking the bullet. Now, what I find actually interesting about this question is the introduction of uncertainty. If you set the odds for either of these, at 95%, or 5%, you can probably get some widely inconsistent and interesting answers.

    Would you take a 5% chance of dying to prevent a 95% chance of democracy collapsing? Would you take a 95% chance of dying to prevent a 5% chance of democracy collapsing? What about 50% and 50%?

    Now, shift those numbers on each side until you get something noteworthy. Personally, I think introducing odds into anything causes humans to lock up and focus on self-preservation. I doubt many people would take their own life to prevent a 40% chance of a family member dying, even if we traditionally valued that family members' life at multiples of our own. This is one of the problems with altruism, and one of the problems with effective donations. Basically, the problem is we don't really know what is going to happen. Even if we are pretty sure our donation money will go to cure someone of a preventable disease, maybe that leads to knock-on effects (they grow up and become bad, overpopulation, the money is stolen and used to buy weapons). Even if the odds of such bad outcomes are extremely low, we become extremely adverse to donating. Maybe I want to donate to AI alignment research, but there is some low probability I make the problem worse. In fact, I really have no idea what will make the problem worse and what will make the problem better. Even if the odds of the money being useful is 80%, that 20% scares me to an irrational level.

    What does this mean? I think it means that research into how effective certain causes are is actually extremely useful. Removing the uncertainty regarding charitable causes might actually be the most impactful contribution of EA, because by doing so we can finally convince people to make a small sacrifice.

Love and Intelligence

    We assume that animals don't really love each other. They partner up out of instinct, and their sexual activities are driven by instinct. Partnership and even family is an animalistic urge, put widely on display across the animal kingdom. Why does a mother lion protect her cubs? Does she actually love them, or is that just an instinct drilled in by millions of years of evolution? These are the thoughts that keep me up at night. Just kidding. But I do wonder if there's actually some sort of spectrum here. If I see a beetle mother taking care of her young, I assume it's all instinct. If I see a female ape taking care of her young, I see more than a glimmer of affection. Maybe this is all personification, but if it is even just a little bit more than that, it is possible that our capacity to love scales directly with either intelligence or sentience.

    Sure, skeptics would just say that love isn't even a real "thing" among humans; it is just another animalistic instinct driven in by evolution, no different than the beetle or fruit fly. Maybe at a certain intelligence level you realize this, and are no longer able to love. A horseshoe theory, where most humans are simply in the sweet spot. Once you pass a certain intelligence threshold or become a certain amount of self-aware, you realize that everything is meaningless and predetermined and love is impossible. Maybe. Maybe love requires a willful suspension of disbelief, but maybe we need to do a better job at separating out lust and sex from this discussion. Love could be an intellectual feat, rather than a physical or spiritual one. Maybe the best marriages and the perfect partnerships could become deeper and more beautiful if each party understood each other on a more fundamental level. Perhaps it is the absence of this understanding that gets in the way of the empathy and compassion really needed for a deeper level of appreciation and respect. I imagine that superintelligent AIs, of the movie Her variety, will be able to form some crazy deep connections. This is based solely on intuition, but it is quite a pleasant thought to believe.

Animal Rights

    Martin McDonagh's film "The Banshees of Inisherin" was my favorite film in 2022. I re-watched it recently, and I was struck by a certain piece of dialogue between one of the main characters, Colm, and a priest:

     Priest: "Do you think God gives a damn about miniature donkeys, Colm?"

    Colm: "I fear he doesn't. And I fear that's where it's all gone wrong."

    Animal rights are a very complex issue, one that I've avoided writing about because I'm not sure exactly where I stand. In Animal Liberation, Peter Singer made a pretty compelling case for vegetarianism. Before we get into what is moral to eat, we first have to solve a hard problem: what makes humans special? In my opinion, there is a flavor of sentience/consciousness that humans have that is extremely valuable. We are by far the most intelligent species, with a level of self awareness and decision making ability far beyond that of other animals. We are the only animal who can contemplate philosophy or wade in existential angst. Our capacity for language and problem solving has allowed us to traverse oceans and explore the stars. Other living creatures on Earth, for the most part, are really dumb. If you spend ten minutes alone with a chicken, you will realize that there is quite simply a not a lot going on upstairs. Yeah we probably shouldn't torture the thing, but if I had to kill a thousand chickens or a random guy, I'd feel pretty confident in my decision. 

    Also, the life of an animal in the wild is not ideal. Evolution is a mostly random process full of cruelty and waste, and most animals die horrifying deaths. Prey are hunted constantly by predators, and predators are constantly at risk of starvation. Starvation, disease, and getting eaten are more than commonplace in the wild, where in farms most animals only have to worry about the last one. Well, maybe some have to worry about living out a miserable life in what is essentially a cramped prison cell where they are kept until slaughter, and that is actually a good point. Here is my personal dilemma: I am confident that I can eat an oyster, and I am confident that we shouldn't eat chimpanzees. Distinct animal species basically either have rights or they don't, and it's weird that society is so logically inconsistent about this (if you eat a dog you are horrible, but go on and kill billions of pigs and that's fine). There is a large spectrum of intelligence from oyster to chimp, and the effective altruists probably draw the line for what we can slaughter too far down. But 97% of humanity draws the line too high, and it's probably better to be safe, especially given how little it costs us. But I find it hard to criticize too harshly whenever I actually see a chicken.

Altruism for the Rich

    Arthur Schopenhauer says that there are two options in life: pain or boredom. The poor deal with pain, the rich deal with boredom. It has been argued that effective altruism is religion for the rich. A way for rich people to feel good about their lives. A way to avoid boredom. A way to donate a small chunk of one's relative wealth and then brag at a dinner party that you "have saved hundreds of lives," all the while treating your employees horribly and scamming customers. This cynical take is fairly common, as many see service work as inherently selfish. Unfortunately, we often miss the obvious. We miss the fact that most people are good, and helping others is not a zero sum game. If helping others makes you feel good, and thus you help others, you are adding value to the world. I don't really care if Bill Gates donates billions of dollars to charity because he truly cares about the world, or if he is donating in order to build a legacy and help his friends. Less children are dying either way. Sure, I hope that everyone eventually migrates to actually altruistic intentions, free of any second-order selfish reasons. But honestly, that is too much to ask in the beginning.

    Social impact is a marathon, not a sprint. If we get bogged down in attacking anyone who tries to do good for "selfish motivations," if we hold everyone to a standard of perfection, we'll lose out on a lot of potential helpers. We miss the forest for the trees, and ignore the fact that the overwhelming majority of people donate essentially nothing. Let's not demand piety, let's just keep each other honest and try to change the world over time. Let's bring people in, not shut people out by gatekeeping. Let's learn from the mistakes of millions of other social movements, and embrace a culture of inclusivity.

Automation Knocks

     A few months ago, I got a letter in the mail. It was a marketing letter, but handwritten on a folded up piece of notebook paper. I thought to myself, wow, I am so much more incentivized to read a marketing letter written with real ink than one of those obviously fake letters written in "handwritten" font. Then I realized that we are in the period of generative AI. AI that can not only come up with customized marketing text for each individual, but AI that can replicate human handwriting via loads of training data. So, if you connected a robot arm to a pen, you could have the arm write very convincing letters in actual ink. These letters would be indistinguishable from human writing, and a collection of robot arms with the same software could write thousands of "handwritten" letters a day. Well, that is quite the business idea. I am sure that millions will be made by the first company to implement such systems, as every marketing guru knows the historic power of the handwritten note. On cue, another layer of human personality will die.

    One of my coworkers asked me for a business idea that he could use for an MBA class project. I looked through my list of notes, and pitched him on this generative AI letter business. It seems to have taken off, with his MBA class now being supremely interested in the idea. The class will be working through the business plan and potentially pitching the idea further up the chain. I might not create the eventual monster, but maybe I pushed it a year or two ahead of schedule.

    Lets take this idea to fruition. In ten years, campaign volunteers are obsolete. Who needs dozens of volunteers to write campaign letters when you can pay a company extremely small sums of money for thousands of equivalent letters? When generative AI starts picking up the phone, grassroots campaigns are essentially dead. The world moves further towards consolidation, with the owners of capital making use of scalable systems to win votes and increase their power. When automation knocks, perhaps we shouldn't answer. Maybe it is the case that when I have a business idea, I should keep it to myself.

The Mediocrity Principle

     Humans often believe that they are at the center of the universe. Disagreeing with this, scientifically speaking, has landed quite a few scientists in trouble. However, essentially all astronomers and cosmologists of the modern day tend to wind up agreeing with The Copernican Principle, which states that the Earth does not occupy a special place in the universe. We are not the center of the universe, and not even the center of our galaxy (gasp!). An extension of this principle is The Mediocrity Principle, the idea that there is nothing special at all about the Earth or humans. In fact, we are probably just par for the course in the universe. We can assume that what is happening here is happening elsewhere (probably even life and intelligent life). This is just a stipulation, but a pretty powerful one. It seems science has trended in this direction, with not just cosmology but also biology (evolution says we are basically just advanced monkeys).

    There is a big problem with this principle: it is quite depressing. We want to think that we are special. We want to strive for important causes and have an impact. We do not want to be forgotten. When we look up at the night sky, it fills us with existential dread to realize that there are more stars in the universe then there are grains of sand on Earth. And a single beach has an incredible amount of sand. The next time you are on a beach, run your fingers through the sand. Imagine the worlds that could exist out if there are really that many stars. Then, wonder if you are really the most important species, the "chosen" race. Probably not. Seems a bit ridiculous. But, maybe? We haven't seen any existence of extraterrestrial life, and the Fermi Paradox is quite complex (why don't we see any evidence for aliens given that we probably should?). Maybe sentient life is extremely rare, or maybe we are the only conscious beings in the universe. This could be important, as without us the universe might just be lifeless balls of gas and rock. We might just be mediocre, or we be the only thing that matters.

    The Mediocrity Principle has rung true up until this point in history. We don't seem particularly special. Whatever we do with nukes, or AI, or pandemics, it doesn't matter much. We could all die out in a flash, and intelligent life elsewhere will probably live on. Perhaps they are better than us, more empathetic and moral. Maybe we would just get in their way. Whatever life force there is in the universe, it doesn't end with us. But, what if it does? Maybe it does? Until there is evidence to the contrary, we have an enormous responsibility. A burden, perhaps. To survive and thrive, to spread among the starts and act as observers to a beautiful universe. Beautiful because of our eyes.

Suffering Risk

    As a longtermist, I care a lot about humanity's potential. I want things to go very well for us: human expansion across the galaxy/universe, free societies full of peace and love, advancements in technology that reduce suffering and death, etc. I do not want things to go very wrong: humanity dies out due to AI/nuclear war/chemically engineered pandemics, authoritarian governments use technology to enslave or repress large swaths of humanity, AI simulates billions of digital minds and puts them in a virtual hellscape, etc. Most people probably agree with this sentiment, they just don't think much about it, and they don't live their lives differently even if they agree. 

    A lot of longtermists care a lot about existential risk, called X-risk for short. Existential risk is the risk that humanity dies out (nukes, pandemics, AI). A different risk is suffering risk, called S-risk. Suffering risk is the risk that humans are stuck in place (authoritarian government takes over and stops progress), or that humans are tortured forever (AI simulates digital minds and tortures them relentlessly, or enslaves humanity and tortures us in "real life"). 80,000 hours estimates that there are less than fifty people in the world who are thinking through how to reduce S-risk. Again, it seems pretty weird that we live in a world of eight billion people where less than fifty are seriously concerned about the very real possibility of worldwide enslavement. For the most part, I don't see any way towards S-risk that isn't agent-driven. What I mean is that only through massive advancements in technology would this sort of grand-scale suffering be possible. We generally think of technology as good, but as Tim Urban writes: “technology is a multiplier of both good and bad. More technology means better good times, but it also means badder bad times.”

    Only through advanced technology do I see such potential for mass suffering. Sure, maybe aliens descend from the sky and torture the species for millennia, but, contrary to what Independence Day would lead you to believe, we probably can't do anything about that. We can, however, massively influence the types of technology that are developed. We can put in place extremely forward-looking safeguards. S-risk is really the main reason I stick so closely to the topic of artificial intelligence. Nukes and pandemics are bad (and I'm sure some authoritarian government could use these to blackmail their citizens), but all you have to do to see S-risk in action is to watch The Matrix one time. Obviously, I doubt robots start using humans for batteries (have they heard about nuclear power?). But in many ways, the paperclip maximizer is the least scary robot.

Lie to Me

     In a perfect world, lawyers do not defend the guilty. We know who the guilty are, and they are adequately punished. This may not be impossible for long. The more data that we collect about an individual, the more we know about them. If you had a camera trained on O. J. Simpson for his entire life, you would know that he was a murderer. If you were a superintelligent AI system, you would likely not have to try that hard to become the world's greatest detective (sorry Batman!) and convict O. J. of murder. Maybe there are actual lie detection techniques that certain AI systems will be good at, but even just by combing through massive amounts of data and using simple inference, I am sure that the policing systems available in the future will be extremely powerful. Powerful enough to trust, and powerful enough to do away with the current "jury of your peers" legal system. Now, there is a trade-off to this, the same trade-off we always face: safety vs human rights. 

    Authoritarian regimes focus on safety. Not safety of their citizens, but safety of their regime. They would want to know when a civilization was lying: "no, I wasn't at the protest last night." They will not want their citizens to have use of this technology: "hey, did you see that Robot 3000 proved that Xi Jinping was lying last night about the Uyghurs?" Use of advanced lie detection in the legal system will certainly change human interaction. Fooling modern-day lie detector machines is a bit of an ironic, because polygraphs are shown to actually not work in any sort of reliable fashion. What about in the future? If your dog knocks over a vase, and then tries to fool you into thinking it happened on its own, would you believe it? We can see right through the attempts of animals to deflect blame or straight up lie. They simply do not have the required mental capacity to string together a convincing argument. That may be us in the future, trying to convince our technocratic overlords of our innocence. It does matter if we turn a 10% wrongfully convicted rate into a 0%, and a 90% rightfully convinced rate into a 100%. What matters much more is what the sentencing requirements are for breaking the law. What matters most is who is writing the laws.

Batman

    From a utilitarian point of view, Batman is the dumbest superhero. After the Joker kills dozens or hundreds of people, Batman tracks him down, beats him up, and sticks him in prison. A prison system that doesn't work, where the Joker always escapes. Then, the Joker kills dozens, hundreds, or billions (in some comics) more. Repeat ad infinitum. This goes for not just the Joker, but for every criminal in Gotham. While this makes for a good storyline (where you don't have to introduce a new bad guy every comic), it is a bit ridiculous. Good thing Batman didn't kill that grunt for stealing a purse, he only pushed him off a building and paralyzed him for life. Now, comic book stans will disagree, with some derivative of "wow it is so cool how morally complex Bruce is, he can't kill because he knows that if he does he won't be able to stop." Other people like The Punisher. Common movie trope: Evil Guy orchestrates a horrific attack on innocent people that kills hundreds. Good Guy fights his way through dozens of low-level bad guys (often brutally killing them) in order to get to the Evil Guy. Good Guy beats Evil Guy and has a chance to kill him. In act act of heroism, Good Guy refuses. Then either the Evil Guy dies anyway or he is locked away (where he often later escapes and kills more people). That is Hollywood, folks.

    I don't want to a world where vigilante justice reigns. In fact, I didn't make this post to lament about superheroes. I wanted to prompt a thought experiment: if you made it your life's mission to decrease the amount of gang violence in the city of Chicago, how much could you accomplish? You could spend every minute calling and emailing politicians, or you could try to run for office. You could work very hard in a high paying profession and donate 50% of your salary to organizations that you believe would decrease crime. What would you do? Think about it for a minute. Whatever it is, I would guess that you would be unlikely to make a big impact. Ken Griffin, billionaire hedge fund founder, couldn't make a dent. How much of an impact could a university make? The University of Chicago, one of the most prestigious universities in the world, can't even protect its students. What could you, personally, hope to accomplish? Without billionaire status or near-superhuman training, you couldn't even be an effective vigilante (Batman or Punisher). Plus, I doubt even an effective vigilante could make dent in a city of almost three million people. Would a real-world Batman fix Chicago? What if he killed people? In both cases, this is a pretty clear "no." Even if Ken Griffin put on the suit and started blasting, I'm pretty sure things would not get any better. Crime is complex. Selling drugs is lucrative. Guns are widely available. Many gang-related shootings are a form of vigilante justice, payback from a previous killing. There are over 100,000 gang members in Chicago. You are just one person. This isn't meant to be a depressing question. Just because your impact might be small doesn't mean that it is worthless. We should still strive for a safer world, regardless of the difficulty. There are no recurring villains in the real world. No easy answers (hey, Batman should just kill people). There are more effective ways to have an impact, but that impact is destined to be small.

The Trade You Make

     There are plenty of existential risks. AI, nuclear war, and chemically engineered pandemics are the most likely and are all pretty horrifying. I was on a run today, and I wondered exactly why I have been stressing so much about these risks. Spending years reading and writing about these topics has made me a more anxious person, hesitant to bring children into a world of suffering on a mass scale. Even if there is only a 10% chance humanity gets wiped out this century, it is clearly possible that the year of this extinction could be a horrible time to endure. I would die, my family would all die, my kids would all die, my wife would die, all of my friends would die, and everyone I've ever met would die. This is very clearly the most depressing outcome imaginable. However, the more I thought about it, the more I realized: this is the trade you make. This is what it means to be human. This is how every human has lived throughout history: in a period of extreme violence and uncertainty.

    If you lived in Germany many hundreds of years ago, you were probably pretty worried that a group of barbarians would raid your cabin, steal your wife, and kill you and your children. If you lived in Egypt thousands of years ago, you were probably a slave and your family could have been killed on a whim by a despot. In fact, in the majority of societies, the majority of people were placed completely at the mercy of authoritarian rulers or warring overlords. The life of a father or mother caught in a current African civil war is par for the course, it is modern life in the West that is the exception. Or is it? Just because we ignore the prospect of nuclear war doesn't mean we don't constantly sit at the brink. Just because we don't read up on chemically engineered pandemics doesn't mean a modified strain of smallpox won't run through New York City and kill 95% of the population. Nothing has changed. Humans have always faced individual and group extinction, we are just moving the boundaries a little more.

    Looking at ex-risk, the only truly scary one from an existential standpoint is AI. Not because it will kill us all (bioweapons and nukes can too), but because it could make us live near-forever in a pit of tremendous suffering. The good news is, this outcome is probably not likely, and the future of AI is basically entirely uncertain. Also, it is the only one out of the three main ex-risks that through development could lead to something resembling a utopia. Regardless, looking too far towards the long term future of humanity can distract from the present. The stresses of current life and the existential struggle we personally have with accepting death are really nothing new. There were plenty of brave men and women who choose to end their lives out of circumstance or out of fear of some circumstance (they heard the barbarians charging over the hill). We should not discount these decisions. But we should also realize that plenty of people stuck it out. Maybe for some pre-programmed survival instinct, or maybe for some other reason. Not only are things pretty good in the world right now on an absolute basis (very, very few people live every day in perpetual bouts of fear and suffering), but things on a relative basis are getting better every year. The long-term trajectory of the human race is positive, perhaps in an exponential fashion. Perhaps not. Personally, I plan to stick around long enough to find out.