TWTS – Episode 25. – Non-Murderous Self Driving Cars
DOWNLOAD (right click to save linked file): TWTS Non-Murderous Self Driving Cars
In response to an article at MIT’s Technology Review magazine that offered the attention-grabbing headline “Why Self Driving Cars Must Be Programmed To Kill”, I feel the need to offer two very crucial points to the discussion:
1. BATNA: The “best alternative to the negotiated agreement” in the operation of a vehicle is to stay centered around a point of health and wellness. All users — be they animal, vegetable, mineral, or otherwise — of a shared space need to direct their actions in such a way as to NOT take chances with anyone’s life, right from the start. The number one rule of sharing spaces with others is to always plan to be able to stop, safely, if something unexpected happens. No plan is ever perfect, but starting out with at least a minimal threshold of something like “Stay safe enough that if there is a physical conflict any potential damage will be fixable”. High risk is only for consensual and limited play, NOT our planet’s basic organism’s infrastructure. Use a truly healthy model for our future systems, not the sick, bad habits of our past or present. In professional conflict resolution and mediation, one of the basic concepts is to always keep in mind that if negotiations are too challenging, one can stop and walk away with nothing worse than the status quo. In other words, no progress is better than making things much worse.
2. Intelligence Increases Options: In a complex system, the ideal plan is NOT a deterministic, direct path from point A to point B, instead the idea plan is an expansive, branching (US American) football-shaped web of as many possible routes to get from here to there as possible, both physically and logistically. A cut-off point that allows for as much wiggle room as possible while still making progress needs to exist before anyone moves forward with sharing the public ways, if we want to have a positive experience as we travel around our world and beyond. Not only do we need wiggle room for other individuals, including inanimate objects, weather, and malfunctioning thingies doing unexpected things, but we also have to include space for confusion and illusion. Senses, and sensors, can be fooled! Making sure that there are as many possible ways out of a potential conflict is the only way to allow for all the possible problems that could happen to happen with as minimal an impact on the goals of everyone sharing the space as possible.
This is just my first reactions, and I could go into far more detail, if anyone wants me to! I just wanted to make it clear that we absolutely do NOT want to program cars to kill, and instead can easily program them to operate in a safe and healthy way?
Links referenced in episode:
More of me!: http://www.reddit.com/user/turil
My other website: http://www.thewiseturtle.com
You can subscribe to The Wise Turtle Speaks Podcast on iTunes or subscribe manually in your favorite podcast application (including iTunes) by putting the rss feed address – https://turil.wordpress.com/category/TWTSpodcast/feed/ – into the “subscribe to podcast” link field (under the file menu in my iTunes).