Robot Servant Problem (excerpt from sci-fi novel Writings in Science)


Loki: I mapped it out on the kitchen table last night. We were trying too hard, that’s why we were stymied.

Nemo: What do you mean?

Loki: We were thinking in terms of an individual being in the machine, as if it needed to do all things for all people, instead of just specific tasks for a specific master. We were mapping out a cat not a dog. That’s the clue. Rethink it as an adjunct to the human, a glove to a hand.

Nemo: Explain.

Loki: Don’t think of the human and robot servant as separate. Think of them as programmer and program. But in this case the programmer is not the human that owns the robot, it’s the human that made the robot. it is he that preprogrammed the robot to fill in the data later and build on that data.

Nemo: Go on.

Loki: The robot doesn’t need to think at all. It first needs to gather info, process info, and react to that info in a pre programmed way.

Nemo: And then?

Loki: We program the robot to monitor the human; his pulse, heart rate, breathing rate, brain waves, eye blinks, blood pressure, etc. Simple stuff. Humans are not that complicated or sophisticated. Their actions are decipherable. It’s easy to tell if they are happy / not happy, angry / not angry, active / not active. We program the robot to monitor the physical changes and respond to each in a specific pre set way.

Nemo: For instance?

Loki: The human becomes active. The robot senses this and responds, “Ready to do something?” or “Going somewhere”, or “What’s up?” Or the human becomes inactive. The robot notices this change and responds with; “Let’s rest,” or “Time for some mental activity.” Or the robot will monitor the signs; hunger signs, sleepy signs, time of day, or in extreme cases, signs of excess anger or fear. In each case he responds as programmed. That’s the basics. Then we take it further.

Nemo: What do you mean?

Loki: After the robot responds, it monitors the human a second time. This time it monitors the response of the human to the robots initial reaction, a feedback loop begins. If the robot response leads to positive reactions in the human, it reinforces that reaction in the robot. If the human reaction leads to a negative reaction, it reinforces in the robot not to repeat that action. Now it’s just simple math. The robot response that leads to a positive response in the owner are repeated or learned. Those that are negative are not. They are deleted or unlearned.

Nemo: So the robot’s job is threefold; monitor the human’s physical reactions, try out preprogrammed responses to those physical reactions, and reinforce those that get positive results while ending those that get negative results, altogether a simple feedback loop.

Loki: Yes! All of this we can do now: monitor human physical behavior, preprogram certain basic responses in the robots we make, and build in a feedback loop to reform that robot response through trial and error, a sort of mechanical selection.

Nemo: We end up with a robot servant with a personality; one that is unconsciously set up by the owner to best cater to his or her specific needs. The robot becomes a glove to the hand.

Loki: A comfortable glove!


Tags: , , ,

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: