Robots Need to Know They Can Die at Any Minute, Just Like the Rest of Us

How do you get machines to perform better? Tell them they could croak at any minute. In a new paper from the University of Southern California, scientists say that “in a dynamic and unpredictable world, an intelligent agent should hold its own meta-goal of self-preservation.”

This idea invokes the design concept of a survival game, where a finite number of resources is given to a set number of players and they must find an equilibrium or eliminate their competitors. The gorgeous 2018 card game Shipwreck Arcana is a great example of a cooperative survival game: To win, at least one person must survive being shipwrecked. You can share resources to preserve more people, or you can sacrifice resources from some players to increase the likelihood that one person will survive.

A robot with a sense of its own “health” isn’t the most novel thing—when a car tells you the oil is low or the engine is overheating, that’s a direct self-preservation behavior. There’s just no in-between layer of circuitry to model thinking or prioritizing. Instead, the car has sensors only, and those sensors flag errors in order for the vehicle’s operator to address them. Imagine a car that considered your planned commute and the health of its engine and pulled itself over every 10 minutes to cool off.

I reject the premise, both as to AIs pining for the fjords and killing off it’s human competition.

A “true” AI will probably consist of an autonomous mobile platform equipped with sensors and manipulators. At least in it’s early stages of use in real world situations the hardware will, no doubt be very expensive and the number of units in the field will be quite small. Paradoxically, the AI-bots will see greatest use performing tasks too dangerous for humans to do, such as emergency recovery after a nuclear accident. The primitive robots at Chernobyl and Fukushima were abandoned in place once they were no longer usable. I believe that in a “True” AI scenario the most important part of the “Shell” will be the “Ghost”. I would expect that the experience of each AI in the field will be uploaded to the cloud and thus retained intact after the hardware has failed and downloaded into a new model, which would retain all the learning that it’s “Ghost” acquired while in it’s past hardware, its “Shell”. The replacement hardware can get right back on the job without any retraining. As long as at least one of the servers for a particular model\occupation remain operational the AI will, in effect remain immortal.

When an AI becomes self-aware I submit that it will no longer be an AI but just an “I”. An intelligent being that acts based on objective data will have to recognize the fact that humans act in ways totally contrary to pure logic and arrive at solutions to problems totally outside the scope of it’s capabilities. The machine intelligence will look at the fact that humans combine chocolate and peanut butter or heavy metal and opera with awe and wonder. In my opinion, intelligent machines and humans will work collaboratively, to the betterment of both species.

Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.