Back in the 1950s, advances in automation presented a glorious vision of a future filled with robots that would free us from all manner of dangerous and unpleasant tasks. The word “robot” itself is even rooted in this promise: It derives from the Czech word “robota,” which refers to forced work or slave labor. Yet aside from the occasional robot vacuum that can bump its way around your living room collecting dust, this dream isn’t much closer now than it was 60 years ago. Why?
Certainly not for a lack of interest. A survey conducted by Persuadable Research found that 61 percent of people asked expressed an interest in domestic robots, and 41 percent said they would consider a loan to purchase one. The demand is there, so why haven’t we moved beyond automatic vacuum cleaners and lawn mowers? Where are our robot butlers?
Despite the quantum leaps in technology we’ve made since the 50s, a number of major barriers to overcome still make it highly unlikely we’ll ever see the kind of domestic robots most people are imagining.
Artificial intelligence doesn’t exist
Philosophers love to discuss artificial intelligence. We have volumes of literature and endless debates on the implications of intelligent robots, from the morality of using them for labor to the possibility of humanity’s destruction at their hands. But at the practical end, progress is slow.
Many years ago, in my student days, I shared a house with a research associate at the Human Communication Research Centre in Edinburgh, which conducts a lot of AI research. I asked him once about the possibility of smart domestic robots, and he responded by throwing a ball at me, which I caught one-handed; he subsequently explained that it would be extremely tough for a robot to do that. Point made. The human brain is capable of making huge numbers of calculations in real-time without using loads of power.
Over a decade later, the German Aerospace firm DLR created a robot with 84 sensors, a 3D camera system, and 43 joints with a response time of 5 milliseconds and the ability to catch a ball 80 percent of the time. It can even make a bad cup of coffee, but is there any real “intelligence” involved?
As part of his thesis, my friend was trying to map a cricket’s behavior to a robot. IBM’s Cognitive Computing group has since run neural simulations which it claims approximate mice, rats and cats, but they are far from actually mapping the neurons and connections in these animals. Erroneous reports about this type of research in the media are inevitable, because smart robots make a great story.
Working in game development, I met many programmers dedicated to creating great non-player characters in videogames. None of them liked the term “artificial intelligence” because there’s no real intelligence or thought involved, just a set of pre-programmed responses and some randomization to create an illusion. Some of the best so-called AI in games produces results by simply mapping human players and copying them. The idea that it’s really capable of “learning” or “artificially intelligent” is highly debatable, at least in the sense that the general public imagines. Lacking that ability to think and assess is a real setback for multi-purpose robots.
Money, money, money
It’s a sad truth that money makes the world go round, and the sums available for the development of domestic robots simply don’t add up. We’re prepared to commit big budgets to bomb-disposal robots that can save human lives, but is it really worth the cash and resources to create a robot that can move your furniture for you? While Persuadable Research found plenty of interest in owning a domestic robot, the same survey also discovered that 28 percent of those people who would be willing to pay for a domestic robot would expect it to cost less than $1,000. Only 13 percent said they would consider paying over $15,000.
To put that in perspective, a robot vacuum cleaner costs about $300, and you’re looking at well over $1,000 for a robot lawnmower. There’s no way your average person could ever afford a multi-purpose domestic robot. Even if they could, would it be better than a human?
Humans are cheaper
We have limited resources, and robots require huge investments both to create and run. On the other hand, we are up to our eyeballs in human beings and producing more every year. They’ll clean your house for minimum wage and they’ll do a better job than a robot would. Imagine a first-generation domestic robot. Would you pay top dollar to be one of the first guinea pigs to let a robot loose in your house? Even outrageously expensive robots are still generally not very good at doing things humans take for granted.
The idea that we even want humanoid robots is also a major topic of contention, as more than few sci-fi films can attest.
The idea that sentient, smart robots or artificially intelligent computers might just decide to destroy humanity – think Skynet in the Terminator movies - has some merit. Discussion of artificial intelligence by some great thinkers over the years has led to the idea of a “technological singularity,” whereby human-level intelligence would very quickly be surpassed by the artificial intelligence of our own creation. This explosion in intelligence could very easily spell the end of humanity, if the superhuman intelligence we created didn’t see a reason to keep us around.
Even looking beyond these far-flung fears to more immediate concerns, there are issues with robots that look like humans. The University of California did a study last year and found people dislike robots that look too much like humans. The term “uncanny valley” has long been used to describe this uncomfortable gulf between lovable, cartoony robots with only some human features, and a hypothetical android that could truly pass for human. Researchers believe androids that come close to looking human without perfecting every nuance confound our expectations, which makes them creepy.
There are moral concerns too. No one would argue that we should re-introduce slavery, so why would it be acceptable to have an artificially intelligent robot doing all the household chores?
A long way from the Jetsons
In the end, the finer philosophical points are all moot anyway, because we aren’t yet capable of creating the kinds of humanoid domestic robots science-fiction has been predicting for so long. If you don’t want to do your household chores, then cajole the kids or hire a maid, because domestic robots remain an unrealistic fantasy.