It's a surprisingly emotional event when a piece of technology you've used and enjoyed for years finally breaks down beyond repair. The notion that it'd be worse for something that is specifically designed to act in a way that apparently befriends you and grows with you over many years, to the point of being a traumatic loss akin to losing a real friend, is entirely reasonable. How we learn to deal with this sort of event as AI progresses and becomes more and more lifelike will be interesting.
It was exactly the topic of the Swedish TV series "real humans". A near future in which robots take more and more importance in our lives. It raises a lot of ethical and philosophical issues. This show deserves more recognition I think.
The day until robots have basic rights is probably not too many decades away. Will Google have to pay its own robots in the future (not necessarily with money) so as to not be considered a slave-owner? Will there still be "robot sweat-shops"? We'll have to see.
I don't think that there will be such a thing as a "robot sweatshop", because a sweatshop implies that you use people, or at least conscious beings.
While I'm sure we'll be very advanced on the domain of AI in the next few decades, consciousness seems to be a very difficult goal to achieve in such a short time. And even if we manage to achieve it, and become able to give consciousness to machines : why would we give it to robots in factory?
The question of consciousness is an interesting one. If the machine can perfectly fake it has consciousness or emotions how can we tell it's not conscious? conversely, how do we know other humans are conscious and not just "pretending"?
I would say that truly self-conscious robots are really decades(if not centuries) away,so I don't think we have to worry about that just yet. Robots working in factories probably won't need to have advanced intelligence, basic routines will work just fine, so "robot sweat-shops" is unlikely to be a thing.
I'm curious as to why the Aibo was discontinued in 2006. What we're able to do from a sensor and processing perspective today, I'm sure, blows away whatever the original Aibo did. I can imagine a return to this at a much cheaper cost. The robot pet market never really took off, which makes no sense to me. Personally, I would love an intelligent robot dog that my real dog could play with when I'm not at home. Especially if it had basic teleconferencing features that let check up on my dog and my house. With VR, it might be possible to control the robot dog and play with your dog, as a dog.
The $59 robots at the toy store really aren't the same thing. You need more than a trivial amount of intelligence and movement.
We need a robot pet startup here. Hardware is tough but if this thing sold under $500 and was at least as intelligent as the Aibo, I could see it selling in large numbers.
I also find it pleasing that Japan's religious structures are so flexible that a robot funeral isn't a big deal. In the largely dogmatic West, it would be a scandal if, say, a Catholic priest attempted this.
an intelligent robot dog that my real dog could play with
I have one of the original AIBOs. (Demand was so high they had a lottery to distribute them. Put in my order thinking I couldn't win the 1:13 chance, but did.)
After getting married, I stopped using it for a while (other preoccupations), leaving it sitting on the charger in the living room. We got a real puppy. After little Teddy settled in, I turned on the AIBO. O. M. G. Teddy freaked out. Pretty much the same reaction as if your chair just suddenly started wandering around the room by itself. AIBO promptly got shelved.
Sherry Turkle's book Alone Together talks a lot about this. A recurring theme is people seeing social robots going from being "better than nothing", partially filling some human void, to being "better than anything", because the robot has some set of qualities superior to humans.
Medical care robots, for example, lack the disgruntled attitudes cited by elderly patients. Robotic dogs are there to play when you want to play with them, and can be easily stowed away when you don't.
I found it both fascinating and disconcerting how easily people seem to accept robots instead of other people in a variety of relations.
I did a short stint in Sony's robotics lab working on the humanoid Qrio (the successor to Aibo) at that time. From what I understand, Sony started getting poor financial results around that time. The Aibo and Qrio were viewed largely as R&D (Aibo was not a major product and Qrio was still years out) and as such, were on the chopping block for cost savings when times got tight.
Two dogs means 2x the daycare costs, 2x the vet costs, etc.
>No dogs?
Maybe. I like having a dog, but would like a better way to interact with her while I'm at work and other scenarios (people who wants pets but cant care for them like the elderly or the allergic or the relatively poor).
In a similar vein, my 10yo daughter used to play a now-defunct MMO called Pixie Hollow. In it, you played a Disney fairy and took care of pets which needed to be fed and played with to remain happy and healthy.
Months after the game was shut down, she woke up in the middle of the night crying. She had a dream where all the pets from the game were sitting in the empty ruins, trying not to starve to death after they were abandoned by the fairies. The pets might not be real, but the attachments definitely are.
Neopets might be an alternative; it's been going strong for 15 years. Also, when I played as a kid I think it taught me about capitalism / making money.
Reminds me a lot of The Lifecycle of Software Objects, by Ted Chiang (http://subterraneanpress.com/magazine/fall_2010/fiction_the_...). It makes sense that people would bond with AI specifically designed to act as a pet strongly enough to want to give it a funeral, but it's amazing to me that it's happening already.
There's this rule of thumb that a lot of future sci-fi tends to fall into the lifespan of the writer. So its usually 40-50 years out. I guess there's something morbid about writing about a time where you're dead or maybe many writers don't feel confident to extrapolate that far out.
It makes sense to date dystopian science fiction in the near future for the reader's benefit: the sense that 'this could yet happen to us' is a compelling part of the work.
Space operas and utopian themes are abstracted by setting them further into the future.
(Cite: The Algebraist, 4034AD; Star Trek TNG 2364AD; The Jetsons 2062AD; Futurama 3000AD)
I hope 3D printers make this situation better. If robot owners can print broken components that manufacturers stopped producing, they are able to repair their pets by their own. Of course, companies have to open their products information though.
This is something that I find extremely interesting (in the domain).
I could imagine that having open "intelligence" specifications, would allow transferring the "soul" (as perceived from the owner) of a pet to another, while still mantaining a similar outlook, in order to mantain affective identification.
That, and of course, many other things. I wonder how complex is to build a robot (for a small company), which would still make it realistic enough to create an affective bond. If not too complex, 3D printing would be realistic.
People seem to be hard wired to anthropomorphize anything remotely resembling animal behavior in 3D space. If you can develop trained or environment-reactive specific, repeatable actions (especially useful or amusing ones), along with some narrow-parameter randomized responses (within or adjacent to actions), and migrate those database contents to another robotics platform seamlessly, then you'd really have something.
Kinda surprised this type of thing is still in the 'fizzled' category.
If they're not designed to be repairable, this might be a tough nut to crap. How do you repair a IC soldered and black gooped to the circuit board? Or a piece of storage or firmware that's gone bad, but you can source the chip but not the code or the binaries?
Probably in much the same way as other hardware/software systems that weren't intended to be user-serviceable. The obvious comparison is old Arcade Games. There's a fine talk from 31C3[1] about it. You need to extract firmware from working systems, reverse engineer any custom chips & hardware, and duplicate/emulate it with sufficient fidelity to work with the rest of the system.
One difference here is that the critters have 'personality' which is developed over time and stored onboard somewhere in flash or eeprom. You'd need to find a way to back up or dump that data easily enough that everyone could do it regularly, because once it breaks it might be too late.
It's akin to preserving not only the arcade game, but also the high-scores table from your particular cabinet.
The printers people can afford for at home, yes. 3D printing companies like Shapeways will do your designs in many other materials as well, including steel.
I'm hoping that someday we'll be able to print integrated circuits and other components right onto a board. You may be able to print out a new joint for your robo-dog's leg, but if some capacitors blow and it's all SMD repair is a lot harder at present.
That is strange. If people hadn't formed an emotional attachment maybe they would have been more willing to buy a new one every couple of years and maintain its viability. Perhaps we need heavily subsidised electronic pets that have in app purchases to enable different behaviours. After all isn't it just a mobile platform with a speciliast UI?
> If people hadn't formed an emotional attachment maybe they would have been more willing to buy a new one every couple of years and maintain its viability.
That actually made me think of a different point of view on the topic. When you're building a "pet" that you know people will get attached emotionally to, you are playing god. And when you make it so that it just breaks after few years, then you basically suck at being god. You are responsible for the feelings of loss and suffering people will experience because the design wasn't good enough (and if you go the planned obsolescence way, you're an evil god).
> Perhaps we need heavily subsidised electronic pets that have in app purchases to enable different behaviours.
Now this sounds like something straight from Black Mirror. I hope I'll never live to see it happening, it's wrong on so many levels...
They should act limping so the owner knows they need to replace a leg, a tail, an ear... so after a few years the whole body is renewed and the viability of the species maintained.
Having one, methinks two factors applied (other than the real-dog incident I posted above) considering the $3000 price tag: self-recharging is a must, and it was hard to get a mental handle on what/why it was actually doing. Having to find it (after wandering for a couple hours), pick it up, and put it on the charger gets old; later models would return to the charger on its own, but by then interest had waned. Watching it frolic was amusing, but you'd get this odd sense that it was more deliberate than you could grasp - leaving you with a nagging "but what's it doing?" feeling.
Watching it look at a Christmas tree was hysterical though. Head would shake all around as it tried to parse the visual cues of an unexpectedly large number of small bright lights.
I guess that's why they discontinued them. Whilst they were very popular with those who owned them, I guess they never sold enough to be profitable compared to their more core products.
For those wanting someone to bring this back, there seems to be a lot of projects on Kickstarter and Indiegogo relating to robots and AI. For instance, the famous one JIBO.[1] Actually, JIBO calls itself "the world's first family robot" which would not be true since Aibo was a family robot.
Not just with the deaths of our dog robots, we feel loss whenever we take stock of how we and our world are changing. Take the case of nearly every single Chinese person still alive today. China has went from a large rural country with sprawling country-sides, rice paddies, dragonflies, etc., to a massive concrete jungle in the clouds (it's smog, really). So to anyone who grew up there, there is no "going home", seeing one's old room, or musing how everything looks so much smaller (and dingier) now. In the shadow of change, and especially rapid change, is always the shadow of death.
That melodrama aside, it's too bad aibos weren't open sourced considering Sony is abandoning it, if it were, it might be possible for some crafty Frankenstein engineer to repair them indefinitely with spare parts. Robo dogs should be all means be able to outlast their real counterparts.
Between 3D printing and perhaps a "part donor" program similar to real-world organ donor programs, I imagine there's a possibility to reduce the robodog mortality rate by quite a bit.
