In problem n.o 2 if we bypass the paradox saying that you don't exist and your grandfather dies, i would not pull the lever as if i pulled it my grandfather dies but that makes it that i didn't exist so the five people also die while letting it go will save me, my siblings, my parents and their siblings and my grandfather in expense of 5 who will die if i pull the lever anyway.
@@c.jishnu378 Only one of your parents would die. And what about if you are an only child? No the real reason why I wouldn't pull it is because of the butterfly effect. You could cause something extremely more catastrophic than simply your grandfather, your parent, and yourself dying. Imagine killing your grandfather would bring about a dictator far worse than Hitler. Sure maybe something bad could happen in the original timeline too, but you wouldn't know that. At the very least, you wouldn't be responsible.
@@centerloper Even if I were the only child, by time travel I wouldn't exist so the trolley will still kill the 5 people meaning 7 will die(atleast), and 5 will die if I did not let the intrusive thoughts win.
it's dumb when it's just a thought experiment, but this has real life implications on artificial intelligence. Also it wouldn't really be fair to call it dumb otherwise either because it's afterall philosophy of ethics.
8:32 It's not a trolley problem, but the best meme I ever saw was of one of those verification things that was a picture of the Ship of Theseus, asking you to select all images which included it lmao
My favorite is an AI asking humans to select all the places they would hide in This one: www.explainxkcd.com/wiki/images/0/01/machine_learning_captcha.png
Interestingly, in germany we got "Eltern haften für ihre Kinder", translating to "Parents are liable for their children", so if your child is 12 and you put it in front of the trolley problem, you _are_ liable for the consequences. It's only at a higher age (14 - 16 I think) where this stops applying
@@jeronimoschujer8060 Not necessarily. I wrote this comment in this context with a purpose. First point: Usually, the legal system builds upon the morality of the many. Therefore it isn't always correct to separate these two. Second point: I kinda agree with how we handle it here. Therefore I am not avoiding the moral problem but rather offering my perspective and my reasoning.
@@nomercy4521 That's kind of my point. I think it does, so stating that your "liable by law" puts the weight on the legal system and avoids deeper (and personal) considerations.
Even in Germany, parent aren't liable for everything their children do, they only have a "duty of supervision" (Aufsichtspflicht) They are only liable for actions done or damages inflicted by their child in case they didn't fulfil this duty Even though there are signs with "parents are liable for their children" all over construction sites and more in Germany, there is no legal basis for that After all, not everything that's written on a sign automatically becomes true just because you put up a sign
AI manages to derail the train and the train runs you over The Ai did this because you built it in this awful universe Did you commit suicide with extra steps?
@@bilboswaggings Suicide would be intentionally causing your own death. It’s very likely you wouldn’t have expected the train to kill you. So instead this might be a freak accident, or the ai might be charged with murder. Depends on if society views artificial intelligence as a sentient being deserving of consequences for their actions or not.
My answer: The people who already died would like to see others not suffer from death, so actually, stopping the trolley as early as possible is justified.
@@legendgames128that completely depends on the type of person already killed, you might feel that way, but others might take comfort in the fact that "all will suffer together" or even be desperate to live that they want to survive even if the others die, and to have the reverse happen would be their worst case scenario You're just assuming the people already dead are maximally virtuous
It's better to argue that, the dead don't have any experience of that unfairness, so you could feasibly save people. But, if you value fairness more, you can just not touch the lever and remain a bystander, which is a huge benefit since you aren't getting involved
Hey Alex, I think the reason why your audience like this video format is because each individual meme whilst being a trolley problem in their own right, open up a doorway into other philosophical matters. You then go on to discuss those and give insight and, with me, expose me to other philosophical dialemma. Keep up the good work!
I think it's a good way to introduce several different philosophical matters in one video. As someone who's going into philosophy in a very casual way (usually as a break from what I do in real life or as something to listen to while I'm doing other things especially chores), I like this kind of videos.
@@alia_babo "Damn, engineering has so much math😮💨... let's ponder insanely specific hypotheticals that I hope & pray to never actually be involved in!"
You missed a crucial piece with the potential infinites: the track that starts with two people has a longer distance before it starts hitting people so at a certain length of "run up" track the first track will have always killed more people at any point in time.
@@frag4007 That's incorrect. Because the above track has +1 person at each comparative group between the top and bottom tracks. So it's an additional +1 to every group. Meaning you've basically added an additional infinity to the top track.
@@Silverizael Depends on how you look at it. If you shift the spot where the people are on the top track one place over, then there are an equal number of people on both tracks, except the bottom track has an extra person at the start. Comparing these kind of series that both tend to infinity is almost always impossible because you can do tricks like these that suddenly make the other series look larger.
@@frag4007 They (will, eventually) have the same number of people overall as people above have already explained, but the point is that this is true only after an infinite amount of time - since the trolley moves at a certain speed instead of simply exterminating all of them simultaneously, so at any given point in time it will have killed a finite number of people. So you can actually compare the killcounts the trolley would make on both of the tracks at any time by looking at when they actually start killing, how much they kill at the beginning and in what manner their scores increase.
They're still around so I'm just commenting to agree that they're dodgy and advise anybody who's thinking about it to look deeper into it before parting with their money.
Anyone who thought they were anything but a gag gift is an idiot who deserves to get scammed. Like who in their right mind thought it was anything other than what it was
It always sounded a but wierd to me, an old law i scotland that if somone own land hes automatically a lord? That would make every free paesant a lord which was not the case the highest % of nobility in european kingdoms was 10% and it was in Poland lithuania under uber specific circumstances, so even if they were actually planting thise trees like they said it still was false advertising
16:25 Oh interesting. When I looked at this, I put the one that starts with 2 people as the lower option/less suffering. I looked at it as having 1 fewer person, as if it had skipped step 1.
AI is not really "it's own intelligence" in the sense that it has free will, it's a model trained to produce an output based on parameters, it's simply another technology, albeit an impressive one. It may be a black box but the box still has boundaries and consequences and we're responsible for how it's used.
Yeah, People don't seem to realize that the AI produces exactly what it was trained on. If it hadn't been trained on space, and avocadoes, it wouldn't be able to create an image of an avocado in space. It's also not creative, the creativity comes from the prompt. It's all just math, very confusing math. But its just math.
@@tobir693but you could argue the human brain is just math as well, since it works in the same way a current computer works just on a different magnitude of complexity
@@tobir693Expanding on what Dark_OFW said, the exact same thing could be said of humans. I couldn't ask you to imagine a color you haven't ever seen-- the limit is set by the cones depicting color in your eyes. Some have fewer than usual, and some have more. But your brain can only imagine what it's been trained on.
@@Jay_Jay_Guy That's not true. We as humans are creative. We can imagine and create things we haven't seen before. Not seeing certain colors is a physical impossibility, not a conceptual one. AI is conceptually limited to what its been trained on, on top of physically. The only weird thing about AI, are emergent features, but I don't know enough about that, to say whether or not that would make AI creative or not. I would lean on no, since they seem to just be features we didn't know we trained the AI on.
In the case of choosing the intensity of pain experienced during one's infinite stay in Hell, there is an easier way to talk about this. Intensity and duration/time are independent variables. Only the duration is infinite. The intensity of pain is constant in either case. Since intensity is the only variable that changes between the two options, the choice depends on that variable.
Ok but hear me out: the top track is the same as the bottom track, just without that first guy, and from that point forward has the same intensity. In this case, you could suffer in hell for all eternity, or suffer in hell for all eternity but you can wait a day before it starts. Like, if you were to put a guy at the beginning of the top track it would literally be the same track, so how could you argue that the sizes are vastly different? The intensity, in this case, is debatable, I think.
@@5eater551 That's incorrect. Because the above track has +1 person at each comparative group between the top and bottom tracks. So it's an additional +1 to every group. Meaning you've basically added an additional infinity to the top track.
@@Silverizael is that the case though? From what we see drawn in the trolley problem, there are three sets of people on the top track before the "…" while there are four on the bottom. So there are two ways to conpare the groups. If we compare them based on their horizontal position then what you said is correct. But if we compare then based on what the "…" is leading us to believe (ie. based on updating both tracks next group of people at the same time) then there is one more person on the bottom track. The latter is also supported by the distance to the first top group being longer than to the first bottom group because of the swerve in the rail. So basically I disagree with his assessment of this problem, since there is no definitive way to interpret the situation. In fact, I also disagree with the analogy he used. Not because it only supports one side of the argument but because I think that the pain you experience over multiple days cannot be summed up for comparison which puts a heavy bias on the choice representing the rail he argues for. Let‘s take the hell analogy and change it up a bit, so that in one option you experience 100 units of pain every day, while in the other one you experience 210 units every other day. Clearly the latter sums up to more units of pain, but I‘m sure there are a lot of people who would still pick it because pain just doesn’t work like that. Also, the analogy left out the core aspect of the problem, which is the stepwise increase of people on the rails, making them very close together at all times. Basically the analogy is biased and not representative of the problem at hand.
@@Silverizael Don't think that's so simple, but dont take my word for it. One thing I'm sure is that you can't simply compute the difference of divergent series without being careful. This is one of those cases where you can write out the sum in different ways and get different results: 1) consider the grouping that you propose: you get for example ( 2 + 3 + 4 + ... ) - ( 1 + 2 + 3 + ...), pairing them like (2-1) + (3-2)+... = 1 + 1 + 1 ... gives infinity, since it's just summing ones. Just like you said ! 2) if you pair the term like (0-1)+(2-2)+(3-3)+... = -1 + 0 + 0 ... = -1. There are infinite terms, so you can always just cancel out with the following, you never get a "leading" term alone if the series is trully infinite We can think of it as: Nº 1) is equivalent to laying down the people like: || ||| |||| | || ||| Nº 2) is equivalent to laying down the people like: || ||| |||| | || ||| |||| The total number of people on the top track didn't change, we just put them a little further away than on the first track. But this makes no sense, your result should not depend on the ordering ! I'm by no means an analysis expert (pretty far from it), so I can't give a sufficiently good answer about this, but there are problems that may arise when you try to compute an infinite sum (additivity and commutation are defined in a binary way, think this is where the issue comes from). If you're curious, googling series rearangments, riemann series theorem, unconditional convergence, and such, should be clarifying. If you're educated on the subject and I'm missing something, let me know !
@@Silverizael That's the same thing. When working with infinities, the counterintuitive fact is that whether you remove the first guy or add an infinite number of people to the track by adding 1 in each position... the total amount is still exactly the same. There's still an infinite number of them, and that infinity is still the same size. However strange we may find it, it changes nothing about the problem. That's just the nature of infinities.
I think the paradox in the fourth case (the one with infinities) is intended a bit different than elaborated here. You can argue that either track is better as follows: 1. the second track is just like the first one, but you miss the first person. After that you kill the same amount every time. So you should choose the second one. 2. the second track is like the first one, but each group of people is one larger than the corresponding group on the first track. So you should go with the first track. It comes down to how you create a correspondence between the people on the tracks (a bijection in math speak). Alex is saying: lets assign each group of people to the group on the other track that is killed at the same point in time. That is one way to do it, and I like the argument with the potential infinite. However, one could argue that this is not clear from the picture and it could as well be that the track lengths are so that you the group with n people is killed after n minutes, no matter the track. In that case Alex' argument should lead to preferring the second track. Also 1=2 only follows from Alex' computation if one is allowed to subtract infinity from both sides, which is not exactly for the reason that this would yield nonsense such as 1=2.
Honestly, this was more of a math-problem than a philosophical one. The answer for me is that it doesn't matter which track you choose, basically the same amount of people are going to die, because stuff gets weird at infinity. lim x -> ∞ = ∞ for both f(x) = x and g(x) = x+1. Every finite number c approaches 0 when compared to infinity. (Example: lim x -> ∞ = 1 for h(x) = (x+c)/x for any real constant c - if I remember my university-math correctly.)
@@romano-britishmedli7407 While I admit that this was my worldview too before I watched the video, I must say now that I found this part addressed very well by Alex. Once you take a limit, you are talking about an actual infinite rather than the potential infinite. Yeah, this is weird philosophy-speak, but I agree that at no point in time infinitely many people are dead, and so it might make sense to compare every point in FINITE time rather than in the limit.
@@TheOneMaddin While I'm a lot into philosophy myself, to be honest I didn't think much about the difference between actual and potential infinity. I just (from a mathematical perspective) found it very unintuitive to stop at a certain point in time, because this would make things finite again (then of course choosing track 2 would be the less bad option, saving one live), and this isn't what the argument is about - for me at least. I'm sorry, but if we were to stop at any point in time (therefore making the whole process finite), why even bring the word "infinity" into the problem? I'm not convinced philosophical definitions are leading to a different conclusion than the mathy one: "It doesn't matter." It sounds to me like "infinity" is a core part of the situaion Alex is presenting (which he can't just remove): "The trolley goes on forever on one of the two tracks, both having infinite people on them." So it's simply nonsensical to "measure" at a certain point in time t-0, because - by definition - it is guaranteed that there will be a next point in time t-1, and then another t-2, and then another t-3..., making thinking about at which point t-x to "stop" the scenario impossible, according to the very description presented. But maybe I understood the problem wrong? Edit: I hope this makes my argument clearer: I guess it depends if the trolley really goes on forever. Because if it does, I can always point to point in time where the second track has caught up with the number of casualties of the first track: Argument: At point t0, there is 1 casualty on track 1 - more than on track 2! Me: "But at point t1, there is also 1 casualty on track 2, *because according to the scenario, the trolley-ride is infinite* ." Argument: But at point tx, there are 300 casualties on track 1 - more than on track 2! Me: "But at point tx+1 there are also 300 casualties on track 2, *because according to the scenario, the trolley-ride is infinite* ." Argument: But at point t?, there are googolplex to the power of googolplex casualties on track 1 - more than on track 2! Me: "But at point t?+1, there are also googolplex to the power of googolplex casualties on track 2, *because according to the scenario, the trolley-ride is infinite* ."
@@romano-britishmedli7407 How about this: the sequence a_n := n is strictly dominated by the sequence b_n := n+1. One can define a partial order on the set of all sequences in this manner. So there is a finer way to compare sequences than to only look at their sum (both sums are infinite). But when you ask whether this matters for this particular problem, then we are of course leaving math behind ...
@@TheOneMaddin Can you explain a little bit more what you mean? It's been some years since I was into university-math. What do you mean by "a finer way to compare sequences"? Honestly, math is the first thing I'd turn towards when someone starts talking about "infinity", since it's such an abstract concept.
I thought you were skeptical enough not to fall for Established Titles. Scottish titles don't work that way; you fell for a scam and are helping scam your audience. This is not an ethical use of your platform.
11:51 The reason we can say the legs and the tabletop are one thing is because the combination of matter serves a definitive purpose. Same with a chair. You could take two saw horses and a door and make a table out of them, and they'd be a table because it serves the purpose of a table.
"the track with two people is going to involve more suffering and death" But what if the trolley, while making the turn, slows down a little bit and is delayed just enough that while the diverted trolley flattens the two people, the non-diverted trolley would also have been flattening the two. Then you've 'saved' one person.
You could also switch to the second track. As the first set of wheels on the trolley start to turn, you have to pull the lever to that the second set of wheels are locked onto the original path. This damages the trolley, but it also means it loses all energy and no one dies.
@@scottthewaterwarrior It doesn't really matter. Multi track drifting is not any more immoral than picking either track for the trolley. If we assume that you divert the first set of wheels but not the second, the trolley probably kills the people at the top a bit earlier. So it'd go something like 2 at the same time, 4, 6, 8, 10 and so on. That's just the even numbers, which is countably infinite, just as the number of people on either track is on its own. If we assume you divert only the second set of wheels, you'd kill 1, 4, 6, 8, 10, and so on. That's also just the even numbers, except 2 got replaced by 1. Still countably infinite. If we assume that the victims on the different tracks are staggered so that the trolley alternately hits _n_ people on one track and then _n_ on the other, it'll go like 1, 2, 2, 3, 3, 4, 4, 5, 5, and so on. This is all the numbers twice, which is still countably infinite. So, I'd say it's worth the gamble. If the trolley stops or derails because of your creative use of the lever, you've saved a countably infinite number of people! And if you accidentally cause it to multi track drift... well, you won't have a countably infinite number of traumatised survivors to accommodate, which arguably is a good thing since overpopulation sucks and you won't need to spread a (presumably) finite amount of resources among them, which would result in no one getting anything, causing a countably infinite degree of suffering. Actually, death might even a mercy if you have finite resources. You won't have needlessly dragged out anyone's suffering, at least. In the end, it's still the same amount of suffering regardless of how long everyone suffered for, but if the trolley causes death quickly enough for there to not be _any_ suffering then you've probably secured the best outcome possible. Now that I've thought it through, I'm actually hoping for the multi-track drifting outcome (again, assuming finite resources to deal with the aftermath). This scenario has already been stretched far beyond what it was meant to make us think about anyway, so I'd say "the fewer survivors, the better" is a perfectly justified conclusion based on what we have to work with. If we _do_ have infinite resources to deal with the aftermath, I guess it might be better to have more survivors, though. In which case multi-track drifting would be the worst outcome, not because it kills the most people (it kills the same amount of people) but because it lets live the least amount of people (which counterintuitively is _not_ strictly tied to how many people are killed). So in that case... just don't touch the lever, I guess.
These trolley problem memes are quite interesting and even fun. I appreciate the exposure to philosophical problems and concepts; your explanations make these issues understandable. Thank you!
I think "AI" is not at the level many people may assume. We may get there. But, at the moment, I'd hold the creator of any computer system responsible for what it does.
Yeah I think so too. I totally disagree with what Alex said and I’m surprised he used “Dali” as an example of how advanced AI is…. Dali is a VERY different form of AI than self driving cars or robots able to make their own moral decisions. Robots are much more programmable than children
@@cubonefan3 Didn’t dig too deep into the topic yet but from what I’ve heard DALL-E is using a classifier to guide the image generation which has been trained on an insane amount of images. I’m hesitant to consider the AI creating entirely new images as he does, if it needs other images to check whether generated noise looks like it’s supposed to. The question is if that’s any different from what humans are doing
@@julianb4333 very different, humans can't interpret exact pixel information the way programs such as dal-e can. These programs are a copyright nightmare because they can be trained using images/photos from artists/photographers without their consent. That is different for the music version of these programs where the developers stated that they won't be using copyrighted music to train them which begs the question why extend that generosity to musicians but not visual artists? Well it may be cause music labels are much quicker with their lawsuit game.
2:40 AI image generation IS "a collage of images that exist on the internet." These algorithms have been fed an enormous amount of images and their descriptions. This is called training data. They've been programmed to identify patterns in those images and descriptions, and then reproduce those patterns on command. The process, very basically, is the programmers going "Okay, algorithm, here's the images that fit these inputs." Followed by "Now, what image would fit THIS input?" What makes the algorithm produce seemingly "new" images is the fact that the algorithm doesn't look directly at the training data, but at the PATTERNS in the training data, with a certain degree of "looseness". This "looseness" is set by the programmer. If it's set too strict, then the algorithm will only be able to reproduce copies of the training data. If it's tooo loose, the algorithm produces gibberish. Get it just right, and it'll spit out something that doesn't look enough like anything in the training data that we assume it's a new thing when, in fact, it's just a remix of a collection of images too large for one human to track, and too dilute for one human to recognize.
I honestly just enjoy listening to Alex talk, whether trolley problems are involved or not. What ever the topic has been, he always seem to have some thoughts on it, that I feel are worth listening to.
So while I DO enjoy the trolley problem videos, for me it was nice to see you dive into philosophical ideas in a low stakes and fun setting. I guess what I’m saying is you don’t have to do meme content for me to enjoy, just fun philosophy where I get to learn something, think, and have fun. Like when you talked about “oh, this problem is a critique of this philosophy and here’s why.” I had fun.
The issue is that if he did a deep enough dive into philosophy he would quickly realize that skepticism isn't a viewpoint that contends with Christianity. The two views don't even interact (and yes I'll call skepticism a view because it is clearly a certain set of characteristics used as a response to various facts).
@@allisthemoist2244 it does though? Christianity, insofar as it is a religion and not just a set of moral guidelines, directly contends with skepticism, as religions are founded on faith, also know as unquestioning belief (unquestioning in regards to the diety existing and being worthy of worship, not in the practices of those governing the faith), whilst skepticism is defined by constant questioning of everything, refusing to accept something based on faith.
@@Santisima_Trinidad you would need to reframe skepticism slightly. The correct definition of it would be "someone who only accepts their cognition as true without evidence" Now, with that said, you would accept justice as real I imagine despite the fact that you don't see it. The reason for this is that you understand justice to be a pattern and you can't see patterns you can only see things that follow them and infer the pattern. In Christianity, we refer to patterns with the word "spirit" which has led to much unfortunate confusion as people wonder why spirits don't just show themselves. Knowing this, we can say that the soul is whatever pattern is unique to you and gives you an identity (the easiest example is a triangle which has three sides as it's soul). That's a brief overview of the concepts of Christianity in a very basic form, but the key thing to notice is that they are only claiming things about the world that everyone already accepts and framing them in a unique way.
@@allisthemoist2244 i can see justice though? Or at the very least, i can see the rules and if they are applied fairly. If they are applied fairly, then that isn't justice. Likewise, if i cannot see the rules, then regardless of concistentcy in judgment then it isn't justice. As for spirit being patterns, you can assign any pattern to being a spirit, but simply because of that we cannot assign a unique spirit to patterns, with any god needing to be a unique spirit for it to be worth believing in as opposed to any other hypothetical spirit. There is no evidence to accept Christianity as truth over anything else, hence it being in opposition to skepticism.
@@Santisima_Trinidad I didn't say we assign a spirit to a pattern. I said we use the word spirit to mean the same thing as the word pattern. And the better way of talking about justice is to say "we can see it in so far as things embody it". or imagine you see tiles layed out red blue green red blue green. In this example, you don't see the pattern but you see things that are the body of the pattern and can tell what it is through them. From those examples, you should see why Catholics talk about spirits like we do, but that it's not contentious claims of the supernatural.
The only reason why you measure a higher number of suffering from the diverted track is because it gets a head start relative to the straight track. If you were to remove the first person on the straight track, then both tracks are equivalent except that the diverted track is shorter before the trollet reaches the victims. If you don't pull the lever, then you have effectively routed the trolley over one extra victim
But it isn't one extra victim, that's the point. For example, imagine you divert the train onto the track that starts with two people. However, before the train gets there, one person from each group manages to escape and run away. Now the diverted track starts with one person instead of two, then two people instead of three, and so on. In other words, it is now identical to the first track. So you started with (supposedly) one less person on the diverted track, then took away an infinite number of people, and ended up with seemingly more than you started with.
That's incorrect. Because the above track has +1 person at each comparative group between the top and bottom tracks. So it's an additional +1 to every group. Meaning you've basically added an additional infinity to the top track.
@@Silverizael Yep. Which is why you can't do ordinary arithmetic with infinities. Each track has the same amount of people, but you can't deduce the amount arithmetically because infinity is not a number, it's a property. Hence why you get stuff like ∞ - 1 = ∞ + ∞.
Nice video Alex. One note on infinities you might want to read up on: in mathematics you actually differentiate sizes of actual infinities. That’s the reason why you can’t just subtract ‘infinity’ from ‘infinity’ - you need to know the substance of each infinity first. For example: (1+2+3+…)/(1x2x3x…) = ‘infinity’/‘infinity’ = 0.
My thought on the ship of Theseus is that as long as there is a planck-time that has one version of you, each version is related to the other and that all these sequential persons have an agreement to take care of each other. If then a past you was reintegrated so that two versions exist at the same time, they will be a new "personality sequence" that is just very similar to you.
12:00 counterpoint, the mug is not fixed to the rest of the table, but the legs and top are fixed to each other. The mug can be removed simply by lifting it off, but it is not so simple for removing the legs from the top.
Thank you for talking about how you only cover a handful of trolley problems. I and others likely would love to see more in a video, but I also bet that I and others like how you open each one up to a discussion or a thought process for philosophical pondering. I like the way you are doing things. I'm excited for a third trolley video!
Please do more trolley problems. It's refreshing to see different versions of this problem, especially when they are combined with other interesting philosophical problems and combined into a very interesting thought experiment!
Something that helps me understand that last point better is Riemann's rearrangement theorem. It's a nice exercise to understand it that I would encourage people to try, and once you do then it really helps understand sequences that tend to infinity better, and also about the rate of divergence. There's a great video by Morphocular on it as well
If the tracks have an infinite number of people, then the relative value of an individual person would be zero. If someone was tied up to a train track and could never escape, they might prefer to get put out of their misery sooner. At any speed of train, there would be people on the infinite tracks suffering starvation, pain, getting cancer (even if they were prevented from dying).
I've never seen the grand-father paradox as a paradox. If you somehow time traveled into the past, you've already severed the flow of causal relationships leading to your existence. From the "perspective" of the universe, successful time travel would look like you spontaneously popped into existence for no reason. So, assuming time travel is even possible, I don’t see why you couldn’t kill your grand-father and continue to exist. You’ve already exited the normal flow of causality by time traveling after all.
For the infinite one, it's interesting that you saw the upper row being "more deadly" because each step had one more person in it. I saw that the top row is the bottom row shifted with 1 less person and so took that one as being the "less deadly" version! This goes to show how important the indexing is when it comes to these "potential infinities". Under that framework, 0+2+3+4+... < 1+2+3+4+... but 2+3+4+5+... > 1+2+3+4+... Even though we've only subtracted 0
For the infinite death variant we could view it two ways One of the tracks starts with one person, while the other starts with two. The one starting with one is the exact same as the other, but with one more which one could argue should mean we should divert the trolley to the track starting with two since the result is the same, but with one extra individual. However since the other track starts with two and we can assume the trolley travels the same speed no matter which track, chosing the track starting with two would mean every moment the trolley runs over someone it kills one more than if it was on the track starting with one. Since it's a potential infinite, I would choose to go for the track starting with one, because it means at any given moment it kills, it kills exactly one less than if you chose the track starting with two, thus reducing overall suffering. Yes they will all go towards infinity, however the amount of suffering at any given moment is still less than if you chose the track starting with two
Appreciate that the sponsorship segment in this video is more cleanly compartmentalized than that of the previous video. It really helps absorb the media and the advertisement better.
"do you beleive in" has always struck me as a silly bit of phraseology, i can only "believe in" myself, the question is, do you believe a god exists, or do you believe that aliens walk among us. and christianity and islam SHOULD be referred to as "the christian mythology" and "the islamic mythology" cos people seem to think all this bullcrap is real.
@@HarryNicNicholas "Believe in X" can mean either "think X exists" or "trust X". I believe in reality (I think reality exists). I believe in my colleagues (I trust my colleagues).
@@cjdennis149 There is an extension to the idea of “trust”, if you are considering early Christianity, where it’s the mutually exclusive choice of “do you believe in Caesar, or do you believe in Jesus”. Both Caesar and Jesus say they are “the son of God, the high priest, the ruler of Earth” (Caesar Augustus had his adopted father Julius Caesar declared a god by the Roman Senate). Belief one way or another necessarily reshapes how you act in that society, as you are choosing whose rules you are going to follow (thus believing that one of them is actually the government of Earth and not the other). To rebel against both of their rules would mean that you believe neither of them is the rightful government of Earth. (Note: for reference, this would be from the standpoint of people who have never stepped outside of the Roman Empire.)
the way I see the Ship of Theseus/table thing is that it's not just linguistic, but the parts were purposely made or selected and are meant to be part of a whole thing. The tabletop was made out of the wood to be the top part of a table and the legs were made to support the tabletop. The mug was made to be it's own thing. For the Ship, the new planks are chosen and put onto the ship to replace the old ones as parts of the ship, thus the one with the new planks would be the Ship of Theseus (imo)
wrapping my head around Alex's conclusion for the infinite number of people problem. If the number of people on the track starting with two people is equal to x, then the number of people on the track starting with 1 person is 1 + x, therefore more people are on that track, so more suffering in theory is on that track
If time travel to the past was real, I can't imagine the grandfather paradox being an issue. The concept of diverging timelines is way more logically consistent in my eyes. Doesn't even need to diverge, traveling to the past could just "erase" the future and it's now just as if you poofed out of thin air into this world. If you were to kill your grandfather and travel to the future, you'd just exist even though your parents don't
Alternatively, bootstrap paradoxes are an appealing resolution. If you travel to the past, it will be because you already had materialized in the past, creating the conditions necessary for you to travel back. If you've already travelled back, then no possible action could prevent you from traveling, as you are participating in events already set in stone.
@@jimijenkins2548 while I think the bootstrap paradox makes way more sense than the grandfather paradox, it too poses an issue. All it takes is 1 person to know something they will or won't do when traveling to the past, and change that detail. "Like, yesterday I met with my future self of today. I'll just go back in time 1 day and during our meeting do something different" And bam there is an issue again.
@@LucyTheBox But for them to know that detail, it has happened. And any attempt to tamper with that happening will inevitably result in its fulfillment.
@@jimijenkins2548 I am aware of the hypothetical, I just don't see it being practically possible. It's way too easy to force a situation that can't realistically be immune to alterations.
For the potential infinites problem, the tracks should be switched to make a more difficult problem. In that case, if you pull the lever, you involve yourself but still in theory kill the same number of people, while not pulling the lever doesn’t involve you but kills one more person if the trolley is stoped at any time.
There are kinda a few possibilities with time travel 1: absolute, paradox occurs and breaks stuff 2: erasure, it erases whatever causes the issue (you) so you were never there to pull the lever 3: realignment, it realigns reality, if you pull the lever it doesn’t really matter, someone else takes your grandfather’s place in some way 4: splitting, incoherent events cause time to split, you are now in a new time line where you never were born, but you still exist as you are from one you were born in 5: prevention, it prevents paradoxes, either you can’t pull the lever, it malfunctions in some way, or the trolley falls off the tracks before it hits your grandfather 6: mess, time is a mess and wishy washy and nihilistic drunkness
You’re wrong on the last one, it could be interpreted as the first infinite, but you remove the guy in front. So you saved someone by choosing the second lane, the people on the lane just die a few seconds earlier than they would have had you chosen the first lane.
Some thoughts on problem 3. 1. Isn't the difference between a mug and a table their functionalities? A table consist of legs and top-part because it needs them to exist, they are necessary for a table. A mug on a table can be removed and not dependant on each other to exist, and therefore, is not part of the table. Rather it can be seen as an accessory of the table. Now, one might say, functionality is made from human mind. While it is true, that is not the point. It's not just "lingualistic boundary". A thing can exist not by looking at it from "material" side (atom side) but its functionality side. 2. I'd like to use modern terms like upgrade: "Ship of Theseus 1.1", meaning the second any part of the ship is changed, it is no longer original, rather a new entity. So, technically, it's not the trolley we diverted. 3. Notice how the moral problem on no.3 is just the same as questioning whether person A or B should get run over. However, by adding new factor, which is the crew replacing parts of the trolley, one can say that we have done our best to prevent casuality. The burden now is on the crew: Why aren't they dismantling it rather than making such paradoxical problem? Idk, this is just some random thought in the morning (it's 2am lol), nice video btw, memes turned to moral study :)
@@chandir7752 Yes they are useful, but there's also an unfounded permanence to them. If I remove a leg from my table, and hit you with that, did I hit you with a mace or with a table leg? I think most people would, as a first intuition, say "table leg" (I asked my roommate, without providing context beforehand, and he did say "table leg".). Yet my table leg is a long, top-heavy piece of wood, and in this hypothetical I'm using it to strike as a person. Why is it still a table leg, and not a mace?
@@jakistam1000 because you stated it was a table leg XD. Why do you say the permanence is unfounded? To me it makes sense since we live In a constantly changing world our worldview reflects that: the table may have been a tree at some point but since someone chopped it and then someone else used the planks to make a table you now call it a "wooden table".
@@BombaJead Okay, I'll admit that the question was a bit leading, but what else was I going to do? Actually unscrew the leg, hit him, and then ask to describe what just happened? I still feel that if I did it, then a person watching the whole thing would say that I hit my roommate with a table leg. But I don't have a solid justification. The fact that we live in a changing world is reflected by language being adapted. So the tree->table example demonstrates sacrificing permanence in order to achieve accuracy, and I applaud that. But the table leg->mace example demonstrates permanence which is, in my opinion, unnecessary in this situation. The function of an object was changed from table leg to mace, and yet it's still referred to as a table leg. (More obvious example would be a baseball bat - people evidently refer to them as baseball bats, even if they become weapons). As I'm writing this, I realize that some permanence is required for practicality. It would be inconvenient to change the description of the table from "chess table" to "monitor table" to "dining table" every time I use it for something else. I guess the language needs to strike a balance between general and specific descriptions, and my personal balance is slightly more towards specific, situation-based descriptions. In general, lack of precision is something that I really dislike in natural language. I accept it as a necessary fact of language, but I still don't like it.
The main reason I’m not sure if I buy the whole functionality argument is 1. There are tables which serve more of a function or even functions that old tables don’t serve yet people still refer to old and new tables as tables 2. If in the future someone says “well a table is something which must be able to hold things on top of itself with a base and legs AS WELL AS GIVE BACK MASSAGES” then a lot of people proceeded to agree with him (let’s say a million or even a billion or half the world) what then? Do we just refer to the old label as being true because we called it as such and appeal to it being the original or do we simply see the use for the new label and go along with it? “A table consists of legs and a top part because it needs them to exist” yeah but according to what and why? If there’s more functionality in adding more rules to the label or changing the label entirely and that’s the basis of labels then did the concept itself exist at all irl or would you stick you foot down to say no any of those changes? What if the mug served a large functionality to a table that we then say it couldn’t possibly be a table without? Lmfao it’s 6am and I’m writing novels rip
Visual and/or speaking problem: the mug seems to contain dark liquid, either it should be all white inside, or you should have addressed that in most languages we distinguish between objects bound together to last (like tables) and Objects in proximity not connected or only connected slightly (like mugs with liquid) and we most often wouldn't call the mug with liquid just "mug" but would say "mug of liquid"
@@CarlosMagnusson07 established titles is a company that sells you a tiny bit of land in Scotland (like 1x1 metre) which they claim technically makes you a lord under Scottish law. This is just completely false, you aren't a lord if you own land in Scotland
Please keep these coming! Your explanations of different infinites is spot on, but I feel like logically not pulling the lever in this situation is the more ethical thing to do, as either way infinite people will die, so I’d rather not be the one to make this happen (even if it’s gonna happen anyway) At least the surviving infinity will have a nice hotel to go back to!
Alex, you might want to reconsider working with established titles. There has been a lot of controversy with them recently. As a gag gift it's possibly okay but they basically have asterisks on every claim they make for what they're even selling. What you actually buy is a piece of paper and really nothing else with it. Please look into it.
A note on the trolly of Thesius: The question of the trolly problem is wether you are acting more or less moral by pulling the lever, and in this instance you give others time to stop the trolly on the way to the second person. Also, in this case there is a very good reason to seperate the old and new trolly. The trolly you diverted isn't the same trolly that will kill the person, since if the people replacing parts on it could simply have removed parts until there is no longer a trolly to kill the person.
i think that there’s a reason why we consider the table and the legs to be one object but not the mug. like, without the legs or the tabletop it wouldn’t be the object we named, it wouldn’t serve its purpose, same thing with the ship, the veils are part of the ship cause the ship wouldn’t work without them. idk just a thought, sorry if it’s not well written, english is not my first language
Exactly what I think. Here there is a conceptual difference, not linguistical as Alex suggested; just because language is clear to distinguish two different things it does not mean that the difference is linguistical and arbitrary.
@@supernukey419 That is correct but not the point. The concept of "ship" does not include sailors, even if the ship needs them to work properly. Like "car" does not include "driver" and "hospital" does not include "doctors". Doctors can work outside hospitals and sailors can be momentarily grounded.
Another reasoning is that things are often referred to as one thing when they are together. In my mind the coffee cup becomes part of the table if you nail it to the table.
@@MilkCalf If I have a puzzle, I can take it apart and move the pieces into arbitrary relative positions and rotations. It's about as un-together as it could be. But it's still this one thing called a puzzle. So that can't explain the difference.
Problem 3 sounds like an insurance scheme. 16:25 Another way of solving this is with hyperreal numbers. The sequence (2,3,4,5,…) approaches an infinite hyperreal number that is greater than the hyperreal that (1,2,3,4,…) approaches. This was so much fun!
About infinite amount of people: on the first track, there is 1+2+3+... people, which is as we know - 1/12 people. On another track, there is double that, so - 1/6. It means that trolley going through both, will remove pain from the universe, but on the 2+4+6 track the pain removed will be greater. So you should go with the 246 track.
Just wanted to point out that at about 10:10 you mention that your body replaces all of its cells every 7 years (according to a study, but the exact number of years doesn't matter and it's more about the idea) - that's actually false; the average age of cells in your body is seven years, but some (like in your stomach) survive for hours, some (like in your muscles and bones) survive for a decade or two, and some (like your neurons and some cells in your eyes) never get replaced and, in fact, don't actually have a maximum lifespan - they just die when you do because they aren't upkept anymore.
For the grandfather paradox trolly problem I would divert it to my grandfather. I believe that because of the paradox, there is no way that he would end up dying. Otherwise I wouldn't be born, thus the train didn't succeed in killing him.
With the infinite tracks, if you line up the tracks to pair off the passenger numbers then you get an interesting result. You can match 2 on one track up with 2, 3 up with 3, and so on. Every number of people ties to the second track is found on the first track. But the first track has one person, whereas the second one doesn't. And so, upon reaching infinity the first track contains the set of all positive integers, and the second track contains the set of all positive integers except 1. So if you pull the lever the trolley will in fact kill one less person than if you didn't. Of course until you reach infinity it will appear that the second track contains an infinite amount more people on it than the first, only on reaching infinity it turns out it has one less person.
That is really interesting. I thought the complete opposite on the potential infinite problem. They both have 2+3+4... but the lower track has +1 person. So I would have said that the bottom one has more suffering and dead people due to that 1 extra person.
Both infinites are countable infinities and therefore the same. It’s still true that taking a measurement means one will be greater than the other but the ultimate logic shows them as both countable infinities.
Another Ship of Theseus reference I love is in Edgar Wright's "The World's End". The protagonist is an alcoholic haunted by his glory days and yearns to return to where he spent them. To the extent where he kept his first ever car he got sold in that town for decades, so he still has that nostalgia. But because its 20 years old he's had to replace every part in the car. If a car is nostalgic, and then you replace every part, is it still nostalgic? Great use of the thought experiment I thought.
16:24 the way I look at it is the track that starts with 2 people starts later because that switch adds distance for the trolley to travel. Because of that, the trolley hits the set of 2 people at the same time on both tracks. If you stop to measure the deaths at any points the bottom track will be 1 higher than the top track.
omg, I just watched the first one days ago. It was one of the most intellectually amusing videos I've watched all month. Now a few days later, this video came out. You read my mind.
Here’s an ethical problem. If a company runs a large scale scam, but does it so that they can help plant trees and protect the environment, did they do a bad thing?
(12:00) I'm pretty sure it has to do with the table being put together. If you flip the table, the mug will fall off, because it's separate, but the table is still one piece. If you remove the legs, it's now 5 separate objects, and no longer a table. (12:40) No, the answer is that the sail is _attached_ to the ship and won't come off easily. If you remove them, they're not part of the ship no more. The sail _belongs_ to the ship, but isn't the ship. You can also attach a person to the ship, and consider the person to be part of the ship.
I like these videos because they're simple and somewhat funny for some reason but also you keep revealing nuanced beliefs in your perspective even though it's just you running ppl over every time
The track beginning with two people is the set containing all combinations of adjacent people greater than 1. The track beginning with one person is the set containing all combinations of adjacent people greater than 1, plus 1 extra person at the beginning. The second track has more people, not less
Alex, haven't you considered playing that one trolley problem game and we will see how you decide on the different problems spot on? It would be great content!
An interesting double trolley problem. A train is on a track heading toward a rail switch, which you could switch, and make the train kill 1 person. If you don't pull the lever the train will travel to a second switch, operated by a stranger. If the stranger pulls the lever the train, no one will be ran over, however if they don't, the train will run over 5 people.
12:45 I don't quite agree. Let's go back to the table with a mug put on top of it example. Sure, we can pull apart a table into a tabletop, and legs, but it woulnd't be a table anymore. Why? Because a table is a table precisely because it can function as a table. Simillarly with a ship. If you break the sails off it will no longer be a ship, because it will no longer serve the ship's function - it will no longer be albe to sail with it. A mug put on top of a table only utilizes table's function, and it doesn't give or take any function from the table (unless it serves as a piece of art, then "a table with a mug on top of it" is an object itself). Similarly, a pressence of a human on a boat does not give or take any functions from it. It just utilizes its purpose - the boat is made to transport humans)
Also, to everybody, that is going to read this. I'm sorry for any technical issues with what I wrote. English is not my native language, and it's pretty hard for me to wrtie a text like this in a nice way. I hope, though, that you understood, what I meant
In regards to the last one, there are 2 ways to look at it, there is the way you mentioned, but also: The 2 tracks don’t have the same number of people, they are both infinite, yes, but the second track has one less person, because the second track contains 2+3+4+… people, which we will call n, the first track has 1+2+3+… or 1+n people, both are infinite, but one is still bigger than the other So by diverting, the total number of people that _have_ died at any given finite moment may be comparatively more, but the total number of people that _will_ die is actually less Of course, the reason answer to that last problem is determining how many people the trolly can run over before stopping, as the trolly has finite momentum and each person will slow it down until it can o longer operate, thus saving the people down the line
(10:00) Isn't that false? As far as I've understood it, it's "on average, a cell it's replaced after 7 years", but that doesn't mean _all_ cells are replaced after 7 years, but that there's been X cell replacements after 7 years where X is the total number of cells. The difference is that one cell can be replaced multiple times and add to X, while another cell is never replaced.
1 thing about the AI problem. People don't seem to get this but just because the AI can learn it doesn't mean that the creator isn't the one responsible. You create the method by which it will learn and you give it the data. In any case you have full control over 1 of those 2 things since if the data is just "the internet" you can't control that but in that case we as a humanity take 50% of the blame for whatever that AI does, good or bad
Okay, I have a good one. You actually, mathematically have a third option: try to derail the trolley in order to save all people on both sides. However, there is a chance that the trolley will have an accident resulting in the deaths of everyone on both sides. Do you try to derail the trolley, or stick to the original problem? And if you do attempt to derail the trolley, how would you be morally accountable for the outcome? Now assume you know the exact probability that the trolley will have an accident and try this again. Do you attempt to derail the trolley at 20% chance of accident? 30%? 40%? 10%? 5%? Now consider: since you'd probably surely be willing to take your chances at something like 1%, what probability range would you take your chances at? Less than the reciprocal of the total of people on both sides, or the side with more people? Exactly that? Less or exactly half that? This will be an interesting one
Maths degree guy (who's admittedly quite rusty with this stuff, so I ask any fellow maths people to correct any mistakes!) here - some infinities are indeed larger than others. The simple way to visualize this is the concept of countability. In layman's terms, countability is the idea that, given any point in that infinite set you could count the number of elements less than it using natural numbers (AKA counting numbers - 1,2,3 etc). This is obviously true for counting numbers and most infinite sets derived from them (like say, the set of natural numbers and their halves: 0.5,1,1.5,2,2.5..., 0.5 is the 1st element, 2.5 is the 5th, 200 is the 400th, 1000000 is the 2000000th and so on - you can always asign a natural number to any element of it). There's sets out there that are too big to be countable, though. The real numbers (so any number you can think of that can be expressed without i, the square root of -1) are uncountably large. In fact, the set of real numbers contained between any two other real numbers is uncountably large. Say you wanted to list all the real numbers between 0 and 1 from lowest to highest. Where in that list is 0.1? Well, it's not the fist number, because 0.01 exists, and then there's 0.001, etc etc. Counting from the "lowest" number in that set all the way to 0.1 (not to speak of all the way to 1!) would take an infinite amount of steps - you can't map it to the counting numbers because the counting numbers aren't large enough, and so it's an uncountably infinite set, larger than the set of counting numbers. This is as opposed to the counting numbers - no matter how high a counting number you name, one can still count to it in a finite amount of steps. (Yes, this is oversimplifying a lot, but I'm trying to make it understandable without higher level maths knowledge, so I'm trying to bring it to stuff most people understand, I'm not going to explain injections and bijections or provide rigorous proof. If anyone's interested in a bit of a better proof for the uncountability of real numbers that's still relatively easy to understand, look up Cantor's Diagonal Argument, it's a satisfyingly intuitive one)
16:30 Wait, now hold on. I was reading the picture as if the track starting with two people just had the first person removed, such that you should pull the lever to save the one person. Now, that's the real paradox.
@@alexkennedy7933 sorry I missed your comment but the company was a scam they didn’t give away any land and that’s also not how lord/lairdship law exactly works also there was no evidence they were actually planting/protecting plants
4:00 Like you said, people are going to be responsible for giving their child- say ptsd, then yeah, that sucks, you suck, BUT not everyone with childhood ptsd becomes a serial killer and thus that's not your fault
Boy oh shuck oh golly oh me oh my oh boy oh mama oh daddy, I sure do love establishing titles. WRONG. That's right... Established Titles? It's a certified Scam-O-Roonie, my friend. One more time for the people in the back. Scam-O-A-O-Rooners, baby.
Artificial intelligence doesn't create complex images out of thin air, it's trained on thousands upon thousands of pieces of art from real human artists. Without human art these images could not exist.
yeah, it kinda peeved me when he explained it as if it where magic, while in fact, very subtly, those AI's ARE just rearranging and combining already existing art
@@tyruskarmesin5418 Im not. Name your price and it'll be done, better than an AI too since if it didn't have any images of a wobbleknocker in it's database it would be impossible for it to create one.
About the table and mug: they are differenciated by function. The table has a funcion of holding objects on an elevated position, while the mug of holding liquids or small objects and being easily carriable by hand. Do not confuse language clarity with conceptual difference.
![BOWKYLION - วิงวอน (ex-change) [Official MV]](http://i.ytimg.com/vi/cLdnZWDaO-w/mqdefault.jpg)
Do not touch the lever under any circumstances! No matter how many people you save. The family of the deceased will sue you! -Your lawyer
So.. better have it on the lane with more people, cause the likelyhood of it hitting the whole family is more likely?
Just compulse the trolley back to the hand
@@Nelsathis kek
I wasn't planning on it, lol.
@@hareecionelson5875 Google translated this to “kill me”.
"professionals taking dumb questions extremely seriously" has got to be one of my favourite genres of youtube video
In problem n.o 2 if we bypass the paradox saying that you don't exist and your grandfather dies, i would not pull the lever as if i pulled it my grandfather dies but that makes it that i didn't exist so the five people also die while letting it go will save me, my siblings, my parents and their siblings and my grandfather in expense of 5 who will die if i pull the lever anyway.
@@c.jishnu378 Only one of your parents would die. And what about if you are an only child? No the real reason why I wouldn't pull it is because of the butterfly effect. You could cause something extremely more catastrophic than simply your grandfather, your parent, and yourself dying. Imagine killing your grandfather would bring about a dictator far worse than Hitler. Sure maybe something bad could happen in the original timeline too, but you wouldn't know that. At the very least, you wouldn't be responsible.
@@centerloper Even if I were the only child, by time travel I wouldn't exist so the trolley will still kill the 5 people meaning 7 will die(atleast), and 5 will die if I did not let the intrusive thoughts win.
They ain’t dumb questions!
it's dumb when it's just a thought experiment, but this has real life implications on artificial intelligence. Also it wouldn't really be fair to call it dumb otherwise either because it's afterall philosophy of ethics.
8:32
It's not a trolley problem, but the best meme I ever saw was of one of those verification things that was a picture of the Ship of Theseus, asking you to select all images which included it lmao
My favorite is an AI asking humans to select all the places they would hide in
This one:
www.explainxkcd.com/wiki/images/0/01/machine_learning_captcha.png
That's a good meme
Haha
*sweats nervously*
Interestingly, in germany we got "Eltern haften für ihre Kinder", translating to "Parents are liable for their children", so if your child is 12 and you put it in front of the trolley problem, you _are_ liable for the consequences. It's only at a higher age (14 - 16 I think) where this stops applying
But that's seems more as a legal perpective, which avoids the moral problem.
@@jeronimoschujer8060 Not necessarily. I wrote this comment in this context with a purpose. First point: Usually, the legal system builds upon the morality of the many. Therefore it isn't always correct to separate these two. Second point: I kinda agree with how we handle it here.
Therefore I am not avoiding the moral problem but rather offering my perspective and my reasoning.
@@jeronimoschujer8060 There's no moral problem making a small under developed child making a decision on who lives and who dies?
@@nomercy4521 That's kind of my point. I think it does, so stating that your "liable by law" puts the weight on the legal system and avoids deeper (and personal) considerations.
Even in Germany, parent aren't liable for everything their children do, they only have a "duty of supervision" (Aufsichtspflicht)
They are only liable for actions done or damages inflicted by their child in case they didn't fulfil this duty
Even though there are signs with "parents are liable for their children" all over construction sites and more in Germany, there is no legal basis for that
After all, not everything that's written on a sign automatically becomes true just because you put up a sign
With enough finesse, you can pull the lever at just the right time to derail the trolley and save everyone. The AI will be particularly adept at this.
AI manages to derail the train and the train runs you over
The Ai did this because you built it in this awful universe
Did you commit suicide with extra steps?
@@bilboswaggings Suicide would be intentionally causing your own death. It’s very likely you wouldn’t have expected the train to kill you. So instead this might be a freak accident, or the ai might be charged with murder. Depends on if society views artificial intelligence as a sentient being deserving of consequences for their actions or not.
If you pull the lever with the at the right time with the right finesse, you can also get multi-track drifting, the AI would be adept at this as well
Yeah I've been bringing this up for years and keep being told "you can't do that".
@@ImperfectVoid8479 how big is this trolley and how many people are on it? What's the speed of the trolley?
Next time you should review "you can stop the trolley at any time but doing so would be unfair to those that already died"
Too real
My answer: The people who already died would like to see others not suffer from death, so actually, stopping the trolley as early as possible is justified.
student loans irl
@@legendgames128that completely depends on the type of person already killed, you might feel that way, but others might take comfort in the fact that "all will suffer together" or even be desperate to live that they want to survive even if the others die, and to have the reverse happen would be their worst case scenario
You're just assuming the people already dead are maximally virtuous
It's better to argue that, the dead don't have any experience of that unfairness, so you could feasibly save people. But, if you value fairness more, you can just not touch the lever and remain a bystander, which is a huge benefit since you aren't getting involved
Hey Alex, I think the reason why your audience like this video format is because each individual meme whilst being a trolley problem in their own right, open up a doorway into other philosophical matters. You then go on to discuss those and give insight and, with me, expose me to other philosophical dialemma.
Keep up the good work!
I love the way it's funny lil memes but with a jumble of different parts of philosophy and stuff
captain obvious to the rescue
@@danfroal8057 haha agree, I think 99.9% of viewers know that, rather than laughing at "memes", they ponder ethical dilemmas
I think it's a good way to introduce several different philosophical matters in one video. As someone who's going into philosophy in a very casual way (usually as a break from what I do in real life or as something to listen to while I'm doing other things especially chores), I like this kind of videos.
@@alia_babo "Damn, engineering has so much math😮💨... let's ponder insanely specific hypotheticals that I hope & pray to never actually be involved in!"
You missed a crucial piece with the potential infinites: the track that starts with two people has a longer distance before it starts hitting people so at a certain length of "run up" track the first track will have always killed more people at any point in time.
That is absolutely correct
Actually the number of people on the upper track is 1 smaller than on the bottom
As long as you dont stop to count
Yeah i was thinking the above track has less people to be exact one less person
@@frag4007 That's incorrect. Because the above track has +1 person at each comparative group between the top and bottom tracks. So it's an additional +1 to every group. Meaning you've basically added an additional infinity to the top track.
@@Silverizael Depends on how you look at it. If you shift the spot where the people are on the top track one place over, then there are an equal number of people on both tracks, except the bottom track has an extra person at the start. Comparing these kind of series that both tend to infinity is almost always impossible because you can do tricks like these that suddenly make the other series look larger.
@@frag4007 They (will, eventually) have the same number of people overall as people above have already explained, but the point is that this is true only after an infinite amount of time - since the trolley moves at a certain speed instead of simply exterminating all of them simultaneously, so at any given point in time it will have killed a finite number of people. So you can actually compare the killcounts the trolley would make on both of the tracks at any time by looking at when they actually start killing, how much they kill at the beginning and in what manner their scores increase.
Established titles is a scam btw Alex, I'd sever ties with them immediately.
They're still around so I'm just commenting to agree that they're dodgy and advise anybody who's thinking about it to look deeper into it before parting with their money.
Yep, seeing these old videos sponsored by them is always sad. Just shows how many people they scammed.
th-cam.com/video/nIjkACB2Xqk/w-d-xo.htmlsi=JqMHOeICFa78OgJy
No it was fucking not
Anyone who thought they were anything but a gag gift is an idiot who deserves to get scammed.
Like who in their right mind thought it was anything other than what it was
It always sounded a but wierd to me, an old law i scotland that if somone own land hes automatically a lord? That would make every free paesant a lord which was not the case the highest % of nobility in european kingdoms was 10% and it was in Poland lithuania under uber specific circumstances, so even if they were actually planting thise trees like they said it still was false advertising
16:25 Oh interesting. When I looked at this, I put the one that starts with 2 people as the lower option/less suffering. I looked at it as having 1 fewer person, as if it had skipped step 1.
I don’t really care about wether I’m responsible for the ia’s actions, all I care about is that it doesn’t start making paper clips
Just in case this comment becomes popular. th-cam.com/video/3mk7NVFz_88/w-d-xo.html
It already has
It minimizes paper clips. Best to wipe out humans so they don’t make any.
@@tyruskarmesin5418 *The Paperclip Minimizer*
"Clip this!"
this channel just feels like an interesting conversation with a friend
That one friend who doesn't talk much,doesnt have lots of friends,and somehow pulls out the most interesting topics on earth
More like with an enemy
AI is not really "it's own intelligence" in the sense that it has free will, it's a model trained to produce an output based on parameters, it's simply another technology, albeit an impressive one. It may be a black box but the box still has boundaries and consequences and we're responsible for how it's used.
Current llm's are what you describe. But a hypothetical AGI would be just as much a moral agent as anyone.
Yeah, People don't seem to realize that the AI produces exactly what it was trained on. If it hadn't been trained on space, and avocadoes, it wouldn't be able to create an image of an avocado in space. It's also not creative, the creativity comes from the prompt. It's all just math, very confusing math. But its just math.
@@tobir693but you could argue the human brain is just math as well, since it works in the same way a current computer works just on a different magnitude of complexity
@@tobir693Expanding on what Dark_OFW said, the exact same thing could be said of humans. I couldn't ask you to imagine a color you haven't ever seen-- the limit is set by the cones depicting color in your eyes. Some have fewer than usual, and some have more. But your brain can only imagine what it's been trained on.
@@Jay_Jay_Guy That's not true. We as humans are creative. We can imagine and create things we haven't seen before. Not seeing certain colors is a physical impossibility, not a conceptual one. AI is conceptually limited to what its been trained on, on top of physically.
The only weird thing about AI, are emergent features, but I don't know enough about that, to say whether or not that would make AI creative or not. I would lean on no, since they seem to just be features we didn't know we trained the AI on.
12:25 The Flying Dutchman resolves this by saying "part of the crew, part of the ship"
Wait this isn't a 2008 genre-defining Deathcore masterpiece, this is ethics!
"dammit jim, i' not a philosopher, i'm artificial!"
You just got trolleyed
TROLOLOLOLOLOLOLOLOLOLOL
DAMN IT
Santa shopping on Amazon and the screen just reads "Ass" is the funniest thing ever
In the case of choosing the intensity of pain experienced during one's infinite stay in Hell, there is an easier way to talk about this. Intensity and duration/time are independent variables. Only the duration is infinite. The intensity of pain is constant in either case. Since intensity is the only variable that changes between the two options, the choice depends on that variable.
Ok but hear me out: the top track is the same as the bottom track, just without that first guy, and from that point forward has the same intensity. In this case, you could suffer in hell for all eternity, or suffer in hell for all eternity but you can wait a day before it starts.
Like, if you were to put a guy at the beginning of the top track it would literally be the same track, so how could you argue that the sizes are vastly different? The intensity, in this case, is debatable, I think.
@@5eater551 That's incorrect. Because the above track has +1 person at each comparative group between the top and bottom tracks. So it's an additional +1 to every group. Meaning you've basically added an additional infinity to the top track.
@@Silverizael is that the case though? From what we see drawn in the trolley problem, there are three sets of people on the top track before the "…" while there are four on the bottom. So there are two ways to conpare the groups. If we compare them based on their horizontal position then what you said is correct. But if we compare then based on what the "…" is leading us to believe (ie. based on updating both tracks next group of people at the same time) then there is one more person on the bottom track. The latter is also supported by the distance to the first top group being longer than to the first bottom group because of the swerve in the rail.
So basically I disagree with his assessment of this problem, since there is no definitive way to interpret the situation. In fact, I also disagree with the analogy he used. Not because it only supports one side of the argument but because I think that the pain you experience over multiple days cannot be summed up for comparison which puts a heavy bias on the choice representing the rail he argues for. Let‘s take the hell analogy and change it up a bit, so that in one option you experience 100 units of pain every day, while in the other one you experience 210 units every other day. Clearly the latter sums up to more units of pain, but I‘m sure there are a lot of people who would still pick it because pain just doesn’t work like that. Also, the analogy left out the core aspect of the problem, which is the stepwise increase of people on the rails, making them very close together at all times. Basically the analogy is biased and not representative of the problem at hand.
@@Silverizael Don't think that's so simple, but dont take my word for it. One thing I'm sure is that you can't simply compute the difference of divergent series without being careful. This is one of those cases where you can write out the sum in different ways and get different results:
1) consider the grouping that you propose: you get for example ( 2 + 3 + 4 + ... ) - ( 1 + 2 + 3 + ...), pairing them like (2-1) + (3-2)+... = 1 + 1 + 1 ... gives infinity, since it's just summing ones. Just like you said !
2) if you pair the term like (0-1)+(2-2)+(3-3)+... = -1 + 0 + 0 ... = -1. There are infinite terms, so you can always just cancel out with the following, you never get a "leading" term alone if the series is trully infinite
We can think of it as:
Nº 1) is equivalent to laying down the people like:
|| ||| ||||
| || |||
Nº 2) is equivalent to laying down the people like:
|| ||| ||||
| || ||| ||||
The total number of people on the top track didn't change, we just put them a little further away than on the first track.
But this makes no sense, your result should not depend on the ordering !
I'm by no means an analysis expert (pretty far from it), so I can't give a sufficiently good answer about this, but there are problems that may arise when you try to compute an infinite sum (additivity and commutation are defined in a binary way, think this is where the issue comes from).
If you're curious, googling series rearangments, riemann series theorem, unconditional convergence, and such, should be clarifying. If you're educated on the subject and I'm missing something, let me know !
@@Silverizael That's the same thing. When working with infinities, the counterintuitive fact is that whether you remove the first guy or add an infinite number of people to the track by adding 1 in each position... the total amount is still exactly the same. There's still an infinite number of them, and that infinity is still the same size. However strange we may find it, it changes nothing about the problem. That's just the nature of infinities.
I think the paradox in the fourth case (the one with infinities) is intended a bit different than elaborated here. You can argue that either track is better as follows:
1. the second track is just like the first one, but you miss the first person. After that you kill the same amount every time. So you should choose the second one.
2. the second track is like the first one, but each group of people is one larger than the corresponding group on the first track. So you should go with the first track.
It comes down to how you create a correspondence between the people on the tracks (a bijection in math speak). Alex is saying: lets assign each group of people to the group on the other track that is killed at the same point in time. That is one way to do it, and I like the argument with the potential infinite. However, one could argue that this is not clear from the picture and it could as well be that the track lengths are so that you the group with n people is killed after n minutes, no matter the track. In that case Alex' argument should lead to preferring the second track.
Also 1=2 only follows from Alex' computation if one is allowed to subtract infinity from both sides, which is not exactly for the reason that this would yield nonsense such as 1=2.
Honestly, this was more of a math-problem than a philosophical one.
The answer for me is that it doesn't matter which track you choose, basically the same amount of people are going to die, because stuff gets weird at infinity.
lim x -> ∞ = ∞ for both f(x) = x and g(x) = x+1.
Every finite number c approaches 0 when compared to infinity.
(Example: lim x -> ∞ = 1 for h(x) = (x+c)/x for any real constant c
- if I remember my university-math correctly.)
@@romano-britishmedli7407 While I admit that this was my worldview too before I watched the video, I must say now that I found this part addressed very well by Alex. Once you take a limit, you are talking about an actual infinite rather than the potential infinite. Yeah, this is weird philosophy-speak, but I agree that at no point in time infinitely many people are dead, and so it might make sense to compare every point in FINITE time rather than in the limit.
@@TheOneMaddin While I'm a lot into philosophy myself, to be honest I didn't think much about the difference between actual and potential infinity. I just (from a mathematical perspective) found it very unintuitive to stop at a certain point in time, because this would make things finite again (then of course choosing track 2 would be the less bad option, saving one live), and this isn't what the argument is about - for me at least.
I'm sorry, but if we were to stop at any point in time (therefore making the whole process finite), why even bring the word "infinity" into the problem? I'm not convinced philosophical definitions are leading to a different conclusion than the mathy one: "It doesn't matter."
It sounds to me like "infinity" is a core part of the situaion Alex is presenting (which he can't just remove): "The trolley goes on forever on one of the two tracks, both having infinite people on them." So it's simply nonsensical to "measure" at a certain point in time t-0, because - by definition - it is guaranteed that there will be a next point in time t-1, and then another t-2, and then another t-3..., making thinking about at which point t-x to "stop" the scenario impossible, according to the very description presented.
But maybe I understood the problem wrong?
Edit: I hope this makes my argument clearer:
I guess it depends if the trolley really goes on forever.
Because if it does, I can always point to point in time where the second track has caught up with the number of casualties of the first track:
Argument: At point t0, there is 1 casualty on track 1 - more than on track 2!
Me: "But at point t1, there is also 1 casualty on track 2, *because according to the scenario, the trolley-ride is infinite* ."
Argument: But at point tx, there are 300 casualties on track 1 - more than on track 2!
Me: "But at point tx+1 there are also 300 casualties on track 2, *because according to the scenario, the trolley-ride is infinite* ."
Argument: But at point t?, there are googolplex to the power of googolplex casualties on track 1 - more than on track 2!
Me: "But at point t?+1, there are also googolplex to the power of googolplex casualties on track 2, *because according to the scenario, the trolley-ride is infinite* ."
@@romano-britishmedli7407 How about this: the sequence a_n := n is strictly dominated by the sequence b_n := n+1. One can define a partial order on the set of all sequences in this manner. So there is a finer way to compare sequences than to only look at their sum (both sums are infinite). But when you ask whether this matters for this particular problem, then we are of course leaving math behind ...
@@TheOneMaddin Can you explain a little bit more what you mean?
It's been some years since I was into university-math.
What do you mean by "a finer way to compare sequences"?
Honestly, math is the first thing I'd turn towards when someone starts talking about "infinity", since it's such an abstract concept.
I thought you were skeptical enough not to fall for Established Titles. Scottish titles don't work that way; you fell for a scam and are helping scam your audience. This is not an ethical use of your platform.
11:51 The reason we can say the legs and the tabletop are one thing is because the combination of matter serves a definitive purpose. Same with a chair. You could take two saw horses and a door and make a table out of them, and they'd be a table because it serves the purpose of a table.
So if I use my chair as a table, would you consider that chair a table. There are some problems with your answer.
"the track with two people is going to involve more suffering and death"
But what if the trolley, while making the turn, slows down a little bit and is delayed just enough that while the diverted trolley flattens the two people, the non-diverted trolley would also have been flattening the two. Then you've 'saved' one person.
You could also switch to the second track. As the first set of wheels on the trolley start to turn, you have to pull the lever to that the second set of wheels are locked onto the original path. This damages the trolley, but it also means it loses all energy and no one dies.
Yes, I assumed that this was the actual dilemma in the problem!
@@waynelu8863 That sounds like it would just make the trolley start multi-track drifting...
@@scottthewaterwarrior It's not that wide. Also, the segments aren't rotatable to that degree
@@scottthewaterwarrior It doesn't really matter. Multi track drifting is not any more immoral than picking either track for the trolley.
If we assume that you divert the first set of wheels but not the second, the trolley probably kills the people at the top a bit earlier. So it'd go something like 2 at the same time, 4, 6, 8, 10 and so on. That's just the even numbers, which is countably infinite, just as the number of people on either track is on its own.
If we assume you divert only the second set of wheels, you'd kill 1, 4, 6, 8, 10, and so on. That's also just the even numbers, except 2 got replaced by 1. Still countably infinite.
If we assume that the victims on the different tracks are staggered so that the trolley alternately hits _n_ people on one track and then _n_ on the other, it'll go like 1, 2, 2, 3, 3, 4, 4, 5, 5, and so on. This is all the numbers twice, which is still countably infinite.
So, I'd say it's worth the gamble. If the trolley stops or derails because of your creative use of the lever, you've saved a countably infinite number of people! And if you accidentally cause it to multi track drift... well, you won't have a countably infinite number of traumatised survivors to accommodate, which arguably is a good thing since overpopulation sucks and you won't need to spread a (presumably) finite amount of resources among them, which would result in no one getting anything, causing a countably infinite degree of suffering.
Actually, death might even a mercy if you have finite resources. You won't have needlessly dragged out anyone's suffering, at least. In the end, it's still the same amount of suffering regardless of how long everyone suffered for, but if the trolley causes death quickly enough for there to not be _any_ suffering then you've probably secured the best outcome possible.
Now that I've thought it through, I'm actually hoping for the multi-track drifting outcome (again, assuming finite resources to deal with the aftermath). This scenario has already been stretched far beyond what it was meant to make us think about anyway, so I'd say "the fewer survivors, the better" is a perfectly justified conclusion based on what we have to work with.
If we _do_ have infinite resources to deal with the aftermath, I guess it might be better to have more survivors, though. In which case multi-track drifting would be the worst outcome, not because it kills the most people (it kills the same amount of people) but because it lets live the least amount of people (which counterintuitively is _not_ strictly tied to how many people are killed). So in that case... just don't touch the lever, I guess.
These trolley problem memes are quite interesting and even fun. I appreciate the exposure to philosophical problems and concepts; your explanations make these issues understandable. Thank you!
I think "AI" is not at the level many people may assume. We may get there. But, at the moment, I'd hold the creator of any computer system responsible for what it does.
Yeah I think so too. I totally disagree with what Alex said and I’m surprised he used “Dali” as an example of how advanced AI is…. Dali is a VERY different form of AI than self driving cars or robots able to make their own moral decisions. Robots are much more programmable than children
@@cubonefan3 Didn’t dig too deep into the topic yet but from what I’ve heard DALL-E is using a classifier to guide the image generation which has been trained on an insane amount of images. I’m hesitant to consider the AI creating entirely new images as he does, if it needs other images to check whether generated noise looks like it’s supposed to. The question is if that’s any different from what humans are doing
@@julianb4333 very different, humans can't interpret exact pixel information the way programs such as dal-e can. These programs are a copyright nightmare because they can be trained using images/photos from artists/photographers without their consent. That is different for the music version of these programs where the developers stated that they won't be using copyrighted music to train them which begs the question why extend that generosity to musicians but not visual artists? Well it may be cause music labels are much quicker with their lawsuit game.
@@BombaJeado.o
2:40
AI image generation IS "a collage of images that exist on the internet."
These algorithms have been fed an enormous amount of images and their descriptions. This is called training data. They've been programmed to identify patterns in those images and descriptions, and then reproduce those patterns on command.
The process, very basically, is the programmers going "Okay, algorithm, here's the images that fit these inputs." Followed by "Now, what image would fit THIS input?"
What makes the algorithm produce seemingly "new" images is the fact that the algorithm doesn't look directly at the training data, but at the PATTERNS in the training data, with a certain degree of "looseness".
This "looseness" is set by the programmer. If it's set too strict, then the algorithm will only be able to reproduce copies of the training data. If it's tooo loose, the algorithm produces gibberish. Get it just right, and it'll spit out something that doesn't look enough like anything in the training data that we assume it's a new thing when, in fact, it's just a remix of a collection of images too large for one human to track, and too dilute for one human to recognize.
I honestly just enjoy listening to Alex talk, whether trolley problems are involved or not. What ever the topic has been, he always seem to have some thoughts on it, that I feel are worth listening to.
i like to try counting the hairs on his face.
So while I DO enjoy the trolley problem videos, for me it was nice to see you dive into philosophical ideas in a low stakes and fun setting. I guess what I’m saying is you don’t have to do meme content for me to enjoy, just fun philosophy where I get to learn something, think, and have fun. Like when you talked about “oh, this problem is a critique of this philosophy and here’s why.” I had fun.
The issue is that if he did a deep enough dive into philosophy he would quickly realize that skepticism isn't a viewpoint that contends with Christianity. The two views don't even interact (and yes I'll call skepticism a view because it is clearly a certain set of characteristics used as a response to various facts).
@@allisthemoist2244 it does though? Christianity, insofar as it is a religion and not just a set of moral guidelines, directly contends with skepticism, as religions are founded on faith, also know as unquestioning belief (unquestioning in regards to the diety existing and being worthy of worship, not in the practices of those governing the faith), whilst skepticism is defined by constant questioning of everything, refusing to accept something based on faith.
@@Santisima_Trinidad you would need to reframe skepticism slightly. The correct definition of it would be "someone who only accepts their cognition as true without evidence"
Now, with that said, you would accept justice as real I imagine despite the fact that you don't see it. The reason for this is that you understand justice to be a pattern and you can't see patterns you can only see things that follow them and infer the pattern. In Christianity, we refer to patterns with the word "spirit" which has led to much unfortunate confusion as people wonder why spirits don't just show themselves. Knowing this, we can say that the soul is whatever pattern is unique to you and gives you an identity (the easiest example is a triangle which has three sides as it's soul).
That's a brief overview of the concepts of Christianity in a very basic form, but the key thing to notice is that they are only claiming things about the world that everyone already accepts and framing them in a unique way.
@@allisthemoist2244 i can see justice though? Or at the very least, i can see the rules and if they are applied fairly. If they are applied fairly, then that isn't justice. Likewise, if i cannot see the rules, then regardless of concistentcy in judgment then it isn't justice.
As for spirit being patterns, you can assign any pattern to being a spirit, but simply because of that we cannot assign a unique spirit to patterns, with any god needing to be a unique spirit for it to be worth believing in as opposed to any other hypothetical spirit. There is no evidence to accept Christianity as truth over anything else, hence it being in opposition to skepticism.
@@Santisima_Trinidad I didn't say we assign a spirit to a pattern. I said we use the word spirit to mean the same thing as the word pattern.
And the better way of talking about justice is to say "we can see it in so far as things embody it". or imagine you see tiles layed out red blue green red blue green. In this example, you don't see the pattern but you see things that are the body of the pattern and can tell what it is through them.
From those examples, you should see why Catholics talk about spirits like we do, but that it's not contentious claims of the supernatural.
The only reason why you measure a higher number of suffering from the diverted track is because it gets a head start relative to the straight track. If you were to remove the first person on the straight track, then both tracks are equivalent except that the diverted track is shorter before the trollet reaches the victims. If you don't pull the lever, then you have effectively routed the trolley over one extra victim
But it isn't one extra victim, that's the point. For example, imagine you divert the train onto the track that starts with two people. However, before the train gets there, one person from each group manages to escape and run away. Now the diverted track starts with one person instead of two, then two people instead of three, and so on. In other words, it is now identical to the first track.
So you started with (supposedly) one less person on the diverted track, then took away an infinite number of people, and ended up with seemingly more than you started with.
That's incorrect. Because the above track has +1 person at each comparative group between the top and bottom tracks. So it's an additional +1 to every group. Meaning you've basically added an additional infinity to the top track.
@@Silverizael Yep. Which is why you can't do ordinary arithmetic with infinities. Each track has the same amount of people, but you can't deduce the amount arithmetically because infinity is not a number, it's a property. Hence why you get stuff like ∞ - 1 = ∞ + ∞.
Nice video Alex. One note on infinities you might want to read up on: in mathematics you actually differentiate sizes of actual infinities. That’s the reason why you can’t just subtract ‘infinity’ from ‘infinity’ - you need to know the substance of each infinity first. For example: (1+2+3+…)/(1x2x3x…) = ‘infinity’/‘infinity’ = 0.
This is described as different orders of infinites starting with Aleph Null. Omega is the ultimate infinity described by Cantor.
My thought on the ship of Theseus is that as long as there is a planck-time that has one version of you, each version is related to the other and that all these sequential persons have an agreement to take care of each other. If then a past you was reintegrated so that two versions exist at the same time, they will be a new "personality sequence" that is just very similar to you.
12:00 counterpoint, the mug is not fixed to the rest of the table, but the legs and top are fixed to each other. The mug can be removed simply by lifting it off, but it is not so simple for removing the legs from the top.
Counterpoint, the roof of a convertible car is not fixed to the car, but is a part of the car
Thank you for talking about how you only cover a handful of trolley problems. I and others likely would love to see more in a video, but I also bet that I and others like how you open each one up to a discussion or a thought process for philosophical pondering. I like the way you are doing things. I'm excited for a third trolley video!
Please do more trolley problems. It's refreshing to see different versions of this problem, especially when they are combined with other interesting philosophical problems and combined into a very interesting thought experiment!
Something that helps me understand that last point better is Riemann's rearrangement theorem. It's a nice exercise to understand it that I would encourage people to try, and once you do then it really helps understand sequences that tend to infinity better, and also about the rate of divergence. There's a great video by Morphocular on it as well
If the tracks have an infinite number of people, then the relative value of an individual person would be zero. If someone was tied up to a train track and could never escape, they might prefer to get put out of their misery sooner. At any speed of train, there would be people on the infinite tracks suffering starvation, pain, getting cancer (even if they were prevented from dying).
I love listening to you ramble about philosophy. And then you combine it with memes - just perfect :D keep it coming 🥳
I've never seen the grand-father paradox as a paradox. If you somehow time traveled into the past, you've already severed the flow of causal relationships leading to your existence. From the "perspective" of the universe, successful time travel would look like you spontaneously popped into existence for no reason. So, assuming time travel is even possible, I don’t see why you couldn’t kill your grand-father and continue to exist. You’ve already exited the normal flow of causality by time traveling after all.
Looking forward to the potentially infinite series of trolly problem memes videos
trolley*
For the infinite one, it's interesting that you saw the upper row being "more deadly" because each step had one more person in it. I saw that the top row is the bottom row shifted with 1 less person and so took that one as being the "less deadly" version!
This goes to show how important the indexing is when it comes to these "potential infinities". Under that framework,
0+2+3+4+... < 1+2+3+4+...
but
2+3+4+5+... > 1+2+3+4+...
Even though we've only subtracted 0
Yes, more!
Would definitely love to see you check out the "Prisoner's Trolley Problemma"
For the infinite death variant we could view it two ways
One of the tracks starts with one person, while the other starts with two. The one starting with one is the exact same as the other, but with one more which one could argue should mean we should divert the trolley to the track starting with two since the result is the same, but with one extra individual.
However since the other track starts with two and we can assume the trolley travels the same speed no matter which track, chosing the track starting with two would mean every moment the trolley runs over someone it kills one more than if it was on the track starting with one. Since it's a potential infinite, I would choose to go for the track starting with one, because it means at any given moment it kills, it kills exactly one less than if you chose the track starting with two, thus reducing overall suffering. Yes they will all go towards infinity, however the amount of suffering at any given moment is still less than if you chose the track starting with two
Yes please do more. My god something about trolley problems just tickle my brain in just the right ways.
Appreciate that the sponsorship segment in this video is more cleanly compartmentalized than that of the previous video. It really helps absorb the media and the advertisement better.
Q: _"Do you believe in the Lord Alex?"_
A: _"I believe in myself."_
*nuff said*
"do you beleive in" has always struck me as a silly bit of phraseology, i can only "believe in" myself, the question is, do you believe a god exists, or do you believe that aliens walk among us. and christianity and islam SHOULD be referred to as "the christian mythology" and "the islamic mythology" cos people seem to think all this bullcrap is real.
@@HarryNicNicholas "Believe in X" can mean either "think X exists" or "trust X". I believe in reality (I think reality exists). I believe in my colleagues (I trust my colleagues).
@@cjdennis149 There is an extension to the idea of “trust”, if you are considering early Christianity, where it’s the mutually exclusive choice of “do you believe in Caesar, or do you believe in Jesus”. Both Caesar and Jesus say they are “the son of God, the high priest, the ruler of Earth” (Caesar Augustus had his adopted father Julius Caesar declared a god by the Roman Senate). Belief one way or another necessarily reshapes how you act in that society, as you are choosing whose rules you are going to follow (thus believing that one of them is actually the government of Earth and not the other). To rebel against both of their rules would mean that you believe neither of them is the rightful government of Earth. (Note: for reference, this would be from the standpoint of people who have never stepped outside of the Roman Empire.)
the way I see the Ship of Theseus/table thing is that it's not just linguistic, but the parts were purposely made or selected and are meant to be part of a whole thing. The tabletop was made out of the wood to be the top part of a table and the legs were made to support the tabletop. The mug was made to be it's own thing. For the Ship, the new planks are chosen and put onto the ship to replace the old ones as parts of the ship, thus the one with the new planks would be the Ship of Theseus (imo)
2:35
Wait… that’s not Amazon
wrapping my head around Alex's conclusion for the infinite number of people problem. If the number of people on the track starting with two people is equal to x, then the number of people on the track starting with 1 person is 1 + x, therefore more people are on that track, so more suffering in theory is on that track
If time travel to the past was real, I can't imagine the grandfather paradox being an issue. The concept of diverging timelines is way more logically consistent in my eyes.
Doesn't even need to diverge, traveling to the past could just "erase" the future and it's now just as if you poofed out of thin air into this world. If you were to kill your grandfather and travel to the future, you'd just exist even though your parents don't
Alternatively, bootstrap paradoxes are an appealing resolution. If you travel to the past, it will be because you already had materialized in the past, creating the conditions necessary for you to travel back. If you've already travelled back, then no possible action could prevent you from traveling, as you are participating in events already set in stone.
@@jimijenkins2548 while I think the bootstrap paradox makes way more sense than the grandfather paradox, it too poses an issue. All it takes is 1 person to know something they will or won't do when traveling to the past, and change that detail.
"Like, yesterday I met with my future self of today. I'll just go back in time 1 day and during our meeting do something different"
And bam there is an issue again.
@@LucyTheBox But for them to know that detail, it has happened. And any attempt to tamper with that happening will inevitably result in its fulfillment.
@@jimijenkins2548 I am aware of the hypothetical, I just don't see it being practically possible. It's way too easy to force a situation that can't realistically be immune to alterations.
For the potential infinites problem, the tracks should be switched to make a more difficult problem. In that case, if you pull the lever, you involve yourself but still in theory kill the same number of people, while not pulling the lever doesn’t involve you but kills one more person if the trolley is stoped at any time.
This is my favorite series by far
There are kinda a few possibilities with time travel
1: absolute, paradox occurs and breaks stuff
2: erasure, it erases whatever causes the issue (you) so you were never there to pull the lever
3: realignment, it realigns reality, if you pull the lever it doesn’t really matter, someone else takes your grandfather’s place in some way
4: splitting, incoherent events cause time to split, you are now in a new time line where you never were born, but you still exist as you are from one you were born in
5: prevention, it prevents paradoxes, either you can’t pull the lever, it malfunctions in some way, or the trolley falls off the tracks before it hits your grandfather
6: mess, time is a mess and wishy washy and nihilistic drunkness
"All infinities are infinite, but some are more infinite than others"
This has to be one of the dumbest quotes ever😂
You’re wrong on the last one, it could be interpreted as the first infinite, but you remove the guy in front. So you saved someone by choosing the second lane, the people on the lane just die a few seconds earlier than they would have had you chosen the first lane.
The ship of Theseus paradox could also be witnessed with the sugababes 😂
I love that in the trolley of Theseus, at no point in the replacement, does anyone try to stop the trolley.
Some thoughts on problem 3.
1. Isn't the difference between a mug and a table their functionalities? A table consist of legs and top-part because it needs them to exist, they are necessary for a table. A mug on a table can be removed and not dependant on each other to exist, and therefore, is not part of the table. Rather it can be seen as an accessory of the table.
Now, one might say, functionality is made from human mind. While it is true, that is not the point. It's not just "lingualistic boundary". A thing can exist not by looking at it from "material" side (atom side) but its functionality side.
2. I'd like to use modern terms like upgrade: "Ship of Theseus 1.1", meaning the second any part of the ship is changed, it is no longer original, rather a new entity. So, technically, it's not the trolley we diverted.
3. Notice how the moral problem on no.3 is just the same as questioning whether person A or B should get run over. However, by adding new factor, which is the crew replacing parts of the trolley, one can say that we have done our best to prevent casuality. The burden now is on the crew: Why aren't they dismantling it rather than making such paradoxical problem?
Idk, this is just some random thought in the morning (it's 2am lol), nice video btw, memes turned to moral study :)
yep. Our labels and boundaries are technically subjective but they're not arbitrary. They are reasonable and very useful.
@@chandir7752 Yes they are useful, but there's also an unfounded permanence to them.
If I remove a leg from my table, and hit you with that, did I hit you with a mace or with a table leg? I think most people would, as a first intuition, say "table leg" (I asked my roommate, without providing context beforehand, and he did say "table leg".). Yet my table leg is a long, top-heavy piece of wood, and in this hypothetical I'm using it to strike as a person. Why is it still a table leg, and not a mace?
@@jakistam1000 because you stated it was a table leg XD. Why do you say the permanence is unfounded? To me it makes sense since we live In a constantly changing world our worldview reflects that: the table may have been a tree at some point but since someone chopped it and then someone else used the planks to make a table you now call it a "wooden table".
@@BombaJead Okay, I'll admit that the question was a bit leading, but what else was I going to do? Actually unscrew the leg, hit him, and then ask to describe what just happened?
I still feel that if I did it, then a person watching the whole thing would say that I hit my roommate with a table leg. But I don't have a solid justification.
The fact that we live in a changing world is reflected by language being adapted. So the tree->table example demonstrates sacrificing permanence in order to achieve accuracy, and I applaud that. But the table leg->mace example demonstrates permanence which is, in my opinion, unnecessary in this situation. The function of an object was changed from table leg to mace, and yet it's still referred to as a table leg. (More obvious example would be a baseball bat - people evidently refer to them as baseball bats, even if they become weapons).
As I'm writing this, I realize that some permanence is required for practicality. It would be inconvenient to change the description of the table from "chess table" to "monitor table" to "dining table" every time I use it for something else. I guess the language needs to strike a balance between general and specific descriptions, and my personal balance is slightly more towards specific, situation-based descriptions.
In general, lack of precision is something that I really dislike in natural language. I accept it as a necessary fact of language, but I still don't like it.
The main reason I’m not sure if I buy the whole functionality argument is
1. There are tables which serve more of a function or even functions that old tables don’t serve yet people still refer to old and new tables as tables
2. If in the future someone says “well a table is something which must be able to hold things on top of itself with a base and legs AS WELL AS GIVE BACK MASSAGES” then a lot of people proceeded to agree with him (let’s say a million or even a billion or half the world) what then? Do we just refer to the old label as being true because we called it as such and appeal to it being the original or do we simply see the use for the new label and go along with it?
“A table consists of legs and a top part because it needs them to exist” yeah but according to what and why? If there’s more functionality in adding more rules to the label or changing the label entirely and that’s the basis of labels then did the concept itself exist at all irl or would you stick you foot down to say no any of those changes? What if the mug served a large functionality to a table that we then say it couldn’t possibly be a table without?
Lmfao it’s 6am and I’m writing novels rip
Visual and/or speaking problem: the mug seems to contain dark liquid, either it should be all white inside, or you should have addressed that in most languages we distinguish between objects bound together to last (like tables) and Objects in proximity not connected or only connected slightly (like mugs with liquid) and we most often wouldn't call the mug with liquid just "mug" but would say "mug of liquid"
Bro was really pushing the established titles scam. I wont forget that.
What is that about?
@@CarlosMagnusson07 established titles is a company that sells you a tiny bit of land in Scotland (like 1x1 metre) which they claim technically makes you a lord under Scottish law. This is just completely false, you aren't a lord if you own land in Scotland
@ thanks
Please keep these coming! Your explanations of different infinites is spot on, but I feel like logically not pulling the lever in this situation is the more ethical thing to do, as either way infinite people will die, so I’d rather not be the one to make this happen (even if it’s gonna happen anyway)
At least the surviving infinity will have a nice hotel to go back to!
Alex, you might want to reconsider working with established titles. There has been a lot of controversy with them recently. As a gag gift it's possibly okay but they basically have asterisks on every claim they make for what they're even selling. What you actually buy is a piece of paper and really nothing else with it. Please look into it.
A note on the trolly of Thesius: The question of the trolly problem is wether you are acting more or less moral by pulling the lever, and in this instance you give others time to stop the trolly on the way to the second person. Also, in this case there is a very good reason to seperate the old and new trolly. The trolly you diverted isn't the same trolly that will kill the person, since if the people replacing parts on it could simply have removed parts until there is no longer a trolly to kill the person.
i think that there’s a reason why we consider the table and the legs to be one object but not the mug. like, without the legs or the tabletop it wouldn’t be the object we named, it wouldn’t serve its purpose, same thing with the ship, the veils are part of the ship cause the ship wouldn’t work without them. idk just a thought, sorry if it’s not well written, english is not my first language
Exactly what I think. Here there is a conceptual difference, not linguistical as Alex suggested; just because language is clear to distinguish two different things it does not mean that the difference is linguistical and arbitrary.
The ship wouldn't work without the sailors either, though
@@supernukey419 That is correct but not the point. The concept of "ship" does not include sailors, even if the ship needs them to work properly.
Like "car" does not include "driver" and "hospital" does not include "doctors".
Doctors can work outside hospitals and sailors can be momentarily grounded.
Another reasoning is that things are often referred to as one thing when they are together. In my mind the coffee cup becomes part of the table if you nail it to the table.
@@MilkCalf If I have a puzzle, I can take it apart and move the pieces into arbitrary relative positions and rotations. It's about as un-together as it could be. But it's still this one thing called a puzzle. So that can't explain the difference.
Problem 3 sounds like an insurance scheme.
16:25 Another way of solving this is with hyperreal numbers. The sequence (2,3,4,5,…) approaches an infinite hyperreal number that is greater than the hyperreal that (1,2,3,4,…) approaches.
This was so much fun!
It's both entertaining and informative, so yeah, keep them coming ! :D
About infinite amount of people: on the first track, there is 1+2+3+... people, which is as we know - 1/12 people. On another track, there is double that, so - 1/6. It means that trolley going through both, will remove pain from the universe, but on the 2+4+6 track the pain removed will be greater. So you should go with the 246 track.
You’re blowing my mind… again!
13:27 “ah yes, dont pull the lever and you save 1/12 of a person!”-some mathematician probably
Just wanted to point out that at about 10:10 you mention that your body replaces all of its cells every 7 years (according to a study, but the exact number of years doesn't matter and it's more about the idea) - that's actually false; the average age of cells in your body is seven years, but some (like in your stomach) survive for hours, some (like in your muscles and bones) survive for a decade or two, and some (like your neurons and some cells in your eyes) never get replaced and, in fact, don't actually have a maximum lifespan - they just die when you do because they aren't upkept anymore.
Yeah I was thinking the same thing
For the grandfather paradox trolly problem I would divert it to my grandfather. I believe that because of the paradox, there is no way that he would end up dying. Otherwise I wouldn't be born, thus the train didn't succeed in killing him.
Bro got sponsored by the Established Titles scam
With the infinite tracks, if you line up the tracks to pair off the passenger numbers then you get an interesting result.
You can match 2 on one track up with 2, 3 up with 3, and so on. Every number of people ties to the second track is found on the first track. But the first track has one person, whereas the second one doesn't. And so, upon reaching infinity the first track contains the set of all positive integers, and the second track contains the set of all positive integers except 1.
So if you pull the lever the trolley will in fact kill one less person than if you didn't. Of course until you reach infinity it will appear that the second track contains an infinite amount more people on it than the first, only on reaching infinity it turns out it has one less person.
you shld definitely make this a series 😂
should*
@@JorgetePanete 💀
That is really interesting. I thought the complete opposite on the potential infinite problem.
They both have 2+3+4... but the lower track has +1 person. So I would have said that the bottom one has more suffering and dead people due to that 1 extra person.
Both infinites are countable infinities and therefore the same. It’s still true that taking a measurement means one will be greater than the other but the ultimate logic shows them as both countable infinities.
Another Ship of Theseus reference I love is in Edgar Wright's "The World's End". The protagonist is an alcoholic haunted by his glory days and yearns to return to where he spent them. To the extent where he kept his first ever car he got sold in that town for decades, so he still has that nostalgia. But because its 20 years old he's had to replace every part in the car. If a car is nostalgic, and then you replace every part, is it still nostalgic? Great use of the thought experiment I thought.
I see people saying that established titles is a fraud/scam. Dunno, but maybe it's true
Watch Scott Shafer's video about Established Titles. He explains pretty clearly why it's a scammy company.
16:24 the way I look at it is the track that starts with 2 people starts later because that switch adds distance for the trolley to travel. Because of that, the trolley hits the set of 2 people at the same time on both tracks. If you stop to measure the deaths at any points the bottom track will be 1 higher than the top track.
You should upload more. I eagerly await for your video.
omg, I just watched the first one days ago. It was one of the most intellectually amusing videos I've watched all month. Now a few days later, this video came out. You read my mind.
Here’s an ethical problem. If a company runs a large scale scam, but does it so that they can help plant trees and protect the environment, did they do a bad thing?
Established titles is a scam in case you didn’t know 😎
(12:00) I'm pretty sure it has to do with the table being put together. If you flip the table, the mug will fall off, because it's separate, but the table is still one piece. If you remove the legs, it's now 5 separate objects, and no longer a table.
(12:40) No, the answer is that the sail is _attached_ to the ship and won't come off easily. If you remove them, they're not part of the ship no more. The sail _belongs_ to the ship, but isn't the ship. You can also attach a person to the ship, and consider the person to be part of the ship.
I like these videos because they're simple and somewhat funny for some reason but also you keep revealing nuanced beliefs in your perspective even though it's just you running ppl over every time
The track beginning with two people is the set containing all combinations of adjacent people greater than 1. The track beginning with one person is the set containing all combinations of adjacent people greater than 1, plus 1 extra person at the beginning. The second track has more people, not less
Alex, haven't you considered playing that one trolley problem game and we will see how you decide on the different problems spot on? It would be great content!
An interesting double trolley problem. A train is on a track heading toward a rail switch, which you could switch, and make the train kill 1 person. If you don't pull the lever the train will travel to a second switch, operated by a stranger. If the stranger pulls the lever the train, no one will be ran over, however if they don't, the train will run over 5 people.
12:45 I don't quite agree. Let's go back to the table with a mug put on top of it example. Sure, we can pull apart a table into a tabletop, and legs, but it woulnd't be a table anymore. Why? Because a table is a table precisely because it can function as a table. Simillarly with a ship. If you break the sails off it will no longer be a ship, because it will no longer serve the ship's function - it will no longer be albe to sail with it. A mug put on top of a table only utilizes table's function, and it doesn't give or take any function from the table (unless it serves as a piece of art, then "a table with a mug on top of it" is an object itself). Similarly, a pressence of a human on a boat does not give or take any functions from it. It just utilizes its purpose - the boat is made to transport humans)
Also, to everybody, that is going to read this. I'm sorry for any technical issues with what I wrote. English is not my native language, and it's pretty hard for me to wrtie a text like this in a nice way. I hope, though, that you understood, what I meant
@@dominikszumski3206you did an amazing job, you even included the art exception.
In regards to the last one, there are 2 ways to look at it, there is the way you mentioned, but also:
The 2 tracks don’t have the same number of people, they are both infinite, yes, but the second track has one less person, because the second track contains 2+3+4+… people, which we will call n, the first track has 1+2+3+… or 1+n people, both are infinite, but one is still bigger than the other
So by diverting, the total number of people that _have_ died at any given finite moment may be comparatively more, but the total number of people that _will_ die is actually less
Of course, the reason answer to that last problem is determining how many people the trolly can run over before stopping, as the trolly has finite momentum and each person will slow it down until it can o longer operate, thus saving the people down the line
(10:00) Isn't that false? As far as I've understood it, it's "on average, a cell it's replaced after 7 years", but that doesn't mean _all_ cells are replaced after 7 years, but that there's been X cell replacements after 7 years where X is the total number of cells. The difference is that one cell can be replaced multiple times and add to X, while another cell is never replaced.
1 thing about the AI problem. People don't seem to get this but just because the AI can learn it doesn't mean that the creator isn't the one responsible. You create the method by which it will learn and you give it the data. In any case you have full control over 1 of those 2 things since if the data is just "the internet" you can't control that but in that case we as a humanity take 50% of the blame for whatever that AI does, good or bad
"... unless u are into pain" 🤣
Okay, I have a good one.
You actually, mathematically have a third option: try to derail the trolley in order to save all people on both sides. However, there is a chance that the trolley will have an accident resulting in the deaths of everyone on both sides. Do you try to derail the trolley, or stick to the original problem? And if you do attempt to derail the trolley, how would you be morally accountable for the outcome?
Now assume you know the exact probability that the trolley will have an accident and try this again. Do you attempt to derail the trolley at 20% chance of accident? 30%? 40%? 10%? 5%?
Now consider: since you'd probably surely be willing to take your chances at something like 1%, what probability range would you take your chances at? Less than the reciprocal of the total of people on both sides, or the side with more people? Exactly that? Less or exactly half that?
This will be an interesting one
Maths degree guy (who's admittedly quite rusty with this stuff, so I ask any fellow maths people to correct any mistakes!) here - some infinities are indeed larger than others. The simple way to visualize this is the concept of countability. In layman's terms, countability is the idea that, given any point in that infinite set you could count the number of elements less than it using natural numbers (AKA counting numbers - 1,2,3 etc). This is obviously true for counting numbers and most infinite sets derived from them (like say, the set of natural numbers and their halves: 0.5,1,1.5,2,2.5..., 0.5 is the 1st element, 2.5 is the 5th, 200 is the 400th, 1000000 is the 2000000th and so on - you can always asign a natural number to any element of it).
There's sets out there that are too big to be countable, though. The real numbers (so any number you can think of that can be expressed without i, the square root of -1) are uncountably large. In fact, the set of real numbers contained between any two other real numbers is uncountably large. Say you wanted to list all the real numbers between 0 and 1 from lowest to highest. Where in that list is 0.1? Well, it's not the fist number, because 0.01 exists, and then there's 0.001, etc etc. Counting from the "lowest" number in that set all the way to 0.1 (not to speak of all the way to 1!) would take an infinite amount of steps - you can't map it to the counting numbers because the counting numbers aren't large enough, and so it's an uncountably infinite set, larger than the set of counting numbers. This is as opposed to the counting numbers - no matter how high a counting number you name, one can still count to it in a finite amount of steps.
(Yes, this is oversimplifying a lot, but I'm trying to make it understandable without higher level maths knowledge, so I'm trying to bring it to stuff most people understand, I'm not going to explain injections and bijections or provide rigorous proof. If anyone's interested in a bit of a better proof for the uncountability of real numbers that's still relatively easy to understand, look up Cantor's Diagonal Argument, it's a satisfyingly intuitive one)
16:30 Wait, now hold on. I was reading the picture as if the track starting with two people just had the first person removed, such that you should pull the lever to save the one person. Now, that's the real paradox.
That sponsor did not age well
Why?
It turned out to be a scam
@@alexkennedy7933 sorry I missed your comment but the company was a scam they didn’t give away any land and that’s also not how lord/lairdship law exactly works also there was no evidence they were actually planting/protecting plants
4:00 Like you said, people are going to be responsible for giving their child- say ptsd, then yeah, that sucks, you suck, BUT not everyone with childhood ptsd becomes a serial killer and thus that's not your fault
Really enjoy these trolley problems. They're all always so creative haha.
12:30 all i got for this is the Pirates of the Caribbean "part of the ship, part of the crew"
Established Titles is apparently a scam, just a heads up
Boy oh shuck oh golly oh me oh my oh boy oh mama oh daddy, I sure do love establishing titles. WRONG. That's right... Established Titles? It's a certified Scam-O-Roonie, my friend. One more time for the people in the back. Scam-O-A-O-Rooners, baby.
Artificial intelligence doesn't create complex images out of thin air, it's trained on thousands upon thousands of pieces of art from real human artists. Without human art these images could not exist.
yeah, it kinda peeved me when he explained it as if it where magic, while in fact, very subtly, those AI's ARE just rearranging and combining already existing art
Could you please draw for me a wobbleknocker? Prior to seeing any images or descriptions of what that is, of course, to avoid plagiarism.
@@tyruskarmesin5418 Of course, you want it as a sketch, lineart, flat color or with shading? The price will depend on that.
@@dokchampa9324 Lol. Kinda feels like dodging the point though.
@@tyruskarmesin5418 Im not. Name your price and it'll be done, better than an AI too since if it didn't have any images of a wobbleknocker in it's database it would be impossible for it to create one.
About the table and mug:
they are differenciated by function.
The table has a funcion of holding objects on an elevated position, while the mug of holding liquids or small objects and being easily carriable by hand.
Do not confuse language clarity with conceptual difference.