Robot Caretakers? What Could Go Wrong?

Caretaker robots are coming. It may not be right away, but a future where we live with a robot that helps us get up off the couch, fetches drinks, and answers the door will be here someday. And that's the future the ACCOMPANY project (Acceptable robotiCs COMPanions for AgeiNg Years) is working toward.

Meet the Care-O-bot 3. It can play games, keep track of occupants' positions in the room, and even recognize visitors. Honestly, what they've accomplished so far is pretty impressive considering the state of the various technologies it employs, like motion tracking and facial recognition. Care-O-bot 3 has a limited set of options from which to choose, but it performs the ones it presents strikingly well — and as a demonstration of the platform’s possibilities, it's inspiring. So what does the next step look like? Or the tenth step?

Maybe we should ask Harcourt Fenton Mudd. Liar, thief, con man, and erstwhile ruler of a planet otherwise populated entirely by androids, Harry Mudd would probably be the first to say that said androids would be ideal for the task we have in mind. They originate from somewhere in the Andromeda galaxy and were built by a race of humanoids who programmed them to be the perfect caretakers — which was what they did right up until their makers' home sun went nova and took their civilization with it, leaving only a few scattered outposts where the crews languished and died. Lacking anyone to serve, the androids meandered around pointlessly on their planet until Mudd arrived and gave them what amounted to a renewed purpose, for better or worse. So what makes them the pinnacle of caregiving technology?

Some of the pros Mudd would waste no time pointing out:

•    They look human. Of course, they're not actually human, but it's easier to relate to something when it looks like a member of your own species, isn't it?

•    They're capable of natural language processing. This is an unbelievably huge benefit: instead of having to rely on a system that tries to guess what you want and presents you with lists of pre-set options, you can just ask (or tell) an android to perform whatever tasks you need done as they come up, whenever you think of them.

•    They can brainstorm and carry out fairly sophisticated plans on their own. Any artificial being capable of thinking through and then flawlessly executing a scheme to commandeer a Constitution-class starship is one I want around to help me when I've fallen and I can't get up.

Or do I? Some of the cons Mudd would probably do his best to avoid mentioning might include:

•    While the androids can certainly come up with and implement intricate plans on their own, their programming, combined with their intelligence, results in some creepy ideas and plots  — namely the takeover of the human race. After all, they could take much better care of us if they were in control, right?

•    Androids don't age. Having an eternally youthful being in the house 24/7 seems like something that would eventually become a constant reminder of my own consistent aging, not to mention the onrushing approach of inevitable death; we'll never know what eventually killed the android makers' outpost crews, but "existential despair" certainly can't be ruled out.

•    No matter how sophisticated they are, Mudd's androids still have computers for brains. They can be affected by observing displays of nonsensical behavior and simple logical paradoxes that freeze them into processing loops of inactivity or overheat their processors, resulting in their destruction (actually, that might be a secret pro). I personally would rather have a caretaker robot that looked nothing like a human. Maybe a bunch of them, actually: an army of small robots designed to do particular tasks that could occasionally join forces to perform more complex operations. What I think I actually want is an exocomp.
Think about it:

•    They can fly. If I need help getting up or support walking from place to place, they can lift me out of my chair and brace me as I walk to the fridge. They could probably even carry me if I was feeling particularly lazy.

•    They can replicate their own tools as needed. Two or three exocomps working together could probably whip up a gourmet meal.

•    They have emergent intelligence, even a noble streak. They seem like they'd be happy helping out around the house without trying to take over the place.

We're a long way from exocomps right now, but keep them in mind the next time you see Care-O-bot 3 in action, or even a Roomba; with technology where it is and where it's going, machines can only get better at making our lives easier.

----

Jon Sung is a contributing writer for XPRIZE and copywriting gun-for-hire to startups and ventures all over the San Francisco Bay area. When not wrangling words for business or pleasure, he serves as the captain of the USS Loma Prieta, the hardest-partying Star Trek fan club in San Francisco.

XPRIZE is an innovation engine. We design and operate prize competitions to address global crises and market failures, and incentivize teams around the world to solve them. Currently, we are operating numerous prizes, including the $30M Google Lunar XPRIZE, challenging privately funded teams to successfully land a robot on the Moon’s surface, and the $10M Qualcomm Tricorder XPRIZE, challenging teams around the world to create a portable, wireless, Star Trek-inspired medical device that allows you to monitor your health and medical conditions anywhere, anytime. The result? Radical innovation that will help us all live long and prosper.

Sign up today to join our mission, be a part of our campaign and win collectibles at: tricorderfederation.org.