There is a correct solution to the trolley problem.

It is this: you design the system with positive train control. You make it so the trolley cannot enter a section where there are people, and you make it so the trolley can be stopped quickly if something goes wrong.

I've been told that this is cheating, a cop-out, or avoiding the "interesting" aspects of the question. I don't think it is. Framing the world as a series of individual can't-win situations and lose-lose decisions _is_ the problem.

Don't do that.

@mattdm @saptaks

@mattdm I'm reminded of a quote attributed to Napoleon, perhaps apocryphal, when he was asked how he would escape some hypothetical no-win military situation: my genius abides in my ability to avoid such situations

@johnmark @mattdm Also Napoleon:

@DamonWakes @mattdm OMG... "do as I say, not as I do?"

That graphic is brutal!

@johnmark @DamonWakes

It is famously one of the best visualisations of... anything.

@mattdm @johnmark @DamonWakes yeah, I never understood why though. Impossible to follow unless you already know the subject.

@johnmark @mattdm Good quote, but how can you choose that over some reference to the Kobayashi Maru?

@mattdm Yeah, if you are solving the trolley problem in real life, someone has already made several fundamental mistakes.

As an example: if a car is barreling towards a group of children and the driver can hit the children and save the occupants of the vehicle or can swerve off a cliff, killing the occupants to save the children, the original mistake was traveling at an unsafe velocity in a place where this scenario could occur.

@mattdm @timb_machine yes. However, I think simple games such as https://www.nationstates.net/ should be mandatory for politicians (and good for most people). So that they understand that every decision brings both positive and negative outcomes.

@aharper @mattdm Even further back, the mistake was not installing bollards or equivalent.

@dalias @aharper @mattdm

Bollards solve a world of ills

@darwinwoodka @aharper @mattdm Sooooo many cases of irresponsible and incompetent drivers killing pedestrians and even ppl minding their own business in their own homes and businesses would just not be a thing if bollards were universal anywhere roads go near people.

@dalias @darwinwoodka @aharper @mattdm While I agree that in many circumstances bollards are great, and we definitely under-use them here in the USA, if your go-to for cars hitting houses and businesses is bollards, you may be over-focusing on a specific solution. In general "design streets for safer outcomes" is right though - just there are lots of other ways to make streets safer that we often don't utilize, and in many cases bollards too are far later than the appropriate intervention.

You might also have more in common with Oscar Newman than you're comfortable with.

@aharper @mattdm If a road admits vehicles hitting pedestrians who are not on the road themselves, and if it does not admit steering to hit an effectively immovable obstacle or another vehicle to avoid hitting a pedestrian who has crossed onto the road, then the road design is critically flawed.

@mattdm

> and you make it so the trolley can be stopped quickly if something goes wrong.

Good that you keep human decision-making in the loop.

I watched a documentary yesterday about "phantom braking", a problem whereby (mandated in the EU) car safety systems error out, pull safety belts ultra-tight, block the steering wheel, and put the car into an emergency break with no means of intervention.. even when in the middle of a busy highway.