The Trolley Problem and Ethics of Driverless Cars
Warren Moore is an ex-apple engineer who started his own software consulting company. Now, he lives a lavish life as a digital nomad in Thailand while dishing out advice about 3D graphics to clients halfway across the globe.
In this episode Warren gives general business advice such as when to branch off and start your own business, technology specific advice such as how to maximize your salary as a software engineer, and digital nomad advice on how to permanently live abroad for pennies on the dollar. Additionally, Warren and I rant about futuristic predictions such as universal basic income, omnipotent artificial intelligence, and virtual reality porn. Some interesting topics we cover include:
0:47 – Digital Nomadism: Working your own hours while experiencing the world
9:40 – Software engineer career progression
30:15 – The digital trolley problem: who will your car save?
37:19 – Technology and automation taking away jobs
41:24 – Pros and cons of universal basic income
53:16 – Amazon and Facebook: Friend or foe
1:05:56 – Exploring god awful startup ideas
1:15:50 – Crazy Job Interview Stories
1:29:19 – When is it time to quit your 9 to 5 and start your own thing
1:42:52 – Are coding bootcamps effective?
1:55:14 – Niching down vs. Branching out: maximizing software engineering salary
2:09:54 – Virtual reality sex with your digital Megan Fox
2:19:35 – Are we living in a simulation?
The full podcast episode:
The audio version of this podcast can be found here:
The most interesting excerpts from this podcast can be found here:
Are We Living in a Simulation
VR Porn – Is it Good or Evil
I got one for you then. I know you've heard before. All right, you're programming the self driving car. You're driving down the road. You got too close to the left and the right. And then you've got a whole gaggle of school children that just randomly crossed the Red, you know, program required to drive off the cliff reply to the kids. Yeah. This
kind of sort of metaphor almost metaphysical reframing of a reframing of a metaphysical trolley. trolley. puzzle. Yeah, it's train problem. Yeah, damsel in distress on the train going to killer not actually read a really good take on this several months ago. And I think it it resolves it pretty pretty tightly. The answer is always break as hard as you can, you know, like, there's, there might be some fallback codification around like trying to minimize loss minimize damage, but a car is going to do in that instance, probably something pretty akin to what a human driver would do, right? You try to avoid the obstacle and you break as hard as you possibly can. And yeah, there's definitely questions around you know, do you it in the the perfect knife's edge case? do you harm the operator? Or do you harm some third party,
that's something I'm not really going to get into. Because I think that smarter people than me have probably thought about this and come to come to better conclusions than I will. But I will say that in practice, the right answer for a self driving car to most of these contrived scenarios is break as hard as you can, yeah, this is pretty much never going to happen. Not that there, you might not be on a road where there is a cliff and a bunch of kids, the middle of the road. But what is much more unlikely, is being able to put all that through an algorithm, the algorithm saying, okay, you have exactly two options. Either you die, or the kids die. I mean, algorithms don't really work like that,
right? And then they don't actually have because they're not syncing organisms, they don't even have an understanding of what these concepts are, again, these are just codification of of what what humans what we suppose humans water should do in these scenarios. This is why companies like Cruz and and Apple and Google and Tesla and so on. And Uber so much time into actually, you know, banking miles on the road, because you just need to have a huge catalogue of scenarios, both common and rare, in order to choose how to do the right thing in any given given scenario. And,
you know, these systems are built by people, and there will always be flaws in them. I think that even currently, if you look at the recent rash of problems with Uber autopilot, and admittedly it works spectacularly well, 99.999%
of the time, but if you look at those issues, we still haven't solved even some of the core fundamental stuff like, you know, you're not, there's not like there's a slippery or narrow shoulder you're gonna fall off into a canyon, when you see a person that you're trying to swerve around. It's literally do you even engage the brakes, when you see somebody pushing a bike across the street, maybe that's at least tables style, then maybe let's dial that in first Ellen Botox, I would also vote yes. But as we saw in practice, it doesn't always happen that way.
You know, it's funny because I start introducing these Felicity concepts to people. And that's how I view them. I view them much more as these interesting philosophical thought exercises when I introduce them to someone like you, someone who's intelligent thought about it, we get into this fun philosophical discussion. But what I found is at least 50% of the time the discussion and a devolving down into Oh, we shouldn't have self driving cars, you know, it's gonna run people over and I tell them, Look, you know, whether or not you can, or cannot program these these philosophical ideas, the self driving car is 1,000%
better at driving the newer where does a magnitude provably I forget how long the Google car drove though getting into a wreck, but I think it was hundreds of thousands if not millions, millions? Yeah, miles, a human can barely go three days without up and hitting something. Yeah.
And causing causing a situation that require somebody else to respond, you know, faster than historically should have to. Yeah, I totally agree. And I think that I think these things are going to get shaken out over time. I think that we're in this really uncovering situation right now, where there are so few self driving cars on the road. And there are so many highly publicized gaffes like, you know, to to automatic cars, drive up to a four way stop and just get deadlocked and just sit there, right. Forever. That's something we learned about an operating system. Yeah, exactly. Yeah, it's sort of it's a real world. And again, no circular waiting, or something like that. Yeah, let's take your locks in the opposite, where you release them, etc. Right? Well, actually, that actually speaks to a point and that is coordination, right? Because not only are not only is one automated car better than human driver, it's also the case that when automated vehicles talk to each other, they become a massively more effective Yeah, well,
yeah. And that's one thing that makes me sad. And actually, the speaks much more about capitalism in general than it does about self driving cars, because it's gonna apply to any industry. But it's just a bummer that we have Google Uber Tesla lift everyone competing to make a self driving car. And every company is slowly getting the first 10% of the problem, the first 20% of the problem and eventually someone's going to hit it and get the person to resent that we get is pool our resources, you know, Marxist society cyl, we could have a self driving car by the end of the week. And I've talked to my friends about this. And they say, Look, that's that's a great idea. But that's that's not how the world works. Nobody's going to pull the resources. The reason the reason that these companies are putting hundreds of millions of dollars into making the self driving car because the first one that gets it gets billions of dollars. Let me ask you this, though. I mean, do you think that if, let's say Samsung, Google, Apple, and whoever else for the big players pool their resources, they could make the perfect phone, but
probably not. So I
think here's here's what I'm saying. I think that there is a lot being fed back to the public domain back to the literature that will in fact act as sort of a rising tide for all these companies. And I think that at this juncture kind of competition that we're seeing is immensely valuable for, for building out these these ecosystems, about self driving cars, there will definitely need to be some kind of country version that comes together and says, once there's some critical mass of automatic cars on the road and says, let's figure out how to get these things talking to each other API, bro, we need a API. Let's start self driving car API. It was probably exists. And I think the cruise is one of those companies that's trying to sort of develop this I could be mistaken. But I think they're trying to develop more or less a middleware platform for this kind of thing that could be licensed and adopted by anybody, any any takers? Right. And I think that's how this eventually goes wide. It's not just necessarily going to be one of these proprietary things. But in fact, it's going to be a company like that, who has logged the same number, same numbers of millions of miles and has built this sort of this flexible self driving vehicle platform. I think that's how it eventually becomes truly ubiquitous. I don't Yeah,
well, it's gonna be awesome. Because it's if we get this dialed in, it's going to be almost like quadrupling our infrastructure system because if we get the network communicating, we can have cars driving at 120 miles an hour with three centimeters in between each bumper, I'm probably exaggerating a little bit but for the most part, it's it's almost like doubling our road space.
Sam is an ambassador for personal growth. When Sam started to take action towards a better life, it wasn’t long before he was hooked faster than Captain Blackbeard’s left hand. Years later, Sam strives to produce change in others similar to the identity level transformation which occurred within himself. His aim is to break fulfillment down into a series of straightforward steps, and introduce it into the life of anyone who is willing to embark on the path of action, education and ownership.