SoylentNews is people
SoylentNews
The future is apparently here. And it's creepier than we ever imagined—even when we were playing around with tethering Teddy Ruxpin to the Internet. A Japanese company called Vinclu ("a company that makes crazy things and supports crazy people") is now taking pre-orders from Japan and the United States for a new interactive, artificial-intelligence driven home automation system. Called Gatebox, the new Internet-of-Things product takes Amazon's Alexa, Google Home, Spike Jonze's film Her , and the "holographic" anime characters of Vocaloid concerts to their unified natural conclusion.
Wait, what?
Gatebox, priced at ¥321,840 (about $2,700 US), is squarely targeted at young lonely salarymen and all brands of anime-obsessed otaku—promising the experience of "living with your favorite character." The size of a home coffee-maker, with a footprint no larger than a sheet of A4 printer paper, the device's main feature is a clear projection tube that displays a computer-animated avatar for the AI's "character." Vinclu apparently is planning multiple possible personalities for Gatebox—which, as part of the device's backstory, is a gateway to the dimension the character lives in.
A company like this could release the first strong AI product (kawaii slave?).
Beginner's definition of "waifu" for the uninitiated.
Update: Another article indicates that "[There's also] HDMI and PC inputs to allow the owner to make their own modifications and create their own characters."
(Score: 2) by Refugee from beyond on Tuesday December 20 2016, @05:54PM
I don’t really see any creepiness, personally. If it had strong AI it would be kinda interesting. Although if it is cloud-based it’s a spy. How far are we from strong AIs again?
Instantly better soylentnews: replace background on article and comment titles with #973131.
(Score: 3, Interesting) by bradley13 on Tuesday December 20 2016, @06:36PM
Maybe not creepy - it doesn't go anywhere near the uncanny valley, after all. But it is pretty sad that some people are glad to have a gadget call them "darling" and tell them it missed them while they were at work. From the demo, I don't have the impression that the dialog is terribly clever; really not much more than an enhanced Eliza, with the addition of responding to specific commands. It will almost certainly get repetitive after the first week.
The whole situation reminds me of the old Simon and Garfunkle song [youtube.com] that goes, in part:
Just a come-on from the whores on Seventh Avenue
I do declare there were times when I was so lonesome
I took some comfort there
There are people lonely enough that they need this. It's still a sad commentary on modern (particularly urben) life...
Everyone is somebody else's weirdo.
(Score: 2) by takyon on Tuesday December 20 2016, @06:57PM
They want more (they would prefer strong AI), but they would settle for far less (preprogrammed/cloud bullshit). As long as it doesn't mean rejection (real life partners).
Although, if you had a strong AI, you would figure it could reject someone unless it had enforced friendliness (an active area of "research"). Maybe old school brainwashing is the key.
[SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
(Score: 2) by Refugee from beyond on Tuesday December 20 2016, @08:25PM
Alternatively, you could try another AI. But then again what happens with an AI which nobody needs?
Instantly better soylentnews: replace background on article and comment titles with #973131.
(Score: 2) by takyon on Tuesday December 20 2016, @08:33PM
It gets sad and starts doing virtual drugs while hanging around with forgotten and neglected Tamagotchis [wikipedia.org].
[SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
(Score: 0) by Anonymous Coward on Wednesday December 21 2016, @03:29AM
Is that it is far easier to buy yourself a friend than it is to acquire one through social interaction for many people. (I did not use the term 'make a friend', because in this context, that would be building your own robot to house your Waifu's personality.)
Society in many parts of the world is not conducive to everyone finding themselves a partner. Maybe some don't want it but many do, and without that, or social interaction, or social services to take care of them (either humanely, or inhumanely.) many people just end up falling through the cracks and dying along in their own shit and piss in their tiny little apartment somewhere. Assuming of course they aren't homeless and end up as just another John Doe passing through the morgue, or local trench graves.
(Score: 2) by takyon on Tuesday December 20 2016, @07:16PM
We're probably not as far away as pessimists believe. I'm guessing there will be a "Chicago pile" moment for a neuromorphic architecture, or AI developments completed in secret because of the obvious benefits (point it at a field of science and watch it discover things in 1% of the time).
Although strong AI could be the appropriate antidote to cloud-based Facebook spies, if you mistreat it *and* let it have an active connection to the Internet, it could easily mess up your life.
See my other posts for more.
[SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
(Score: 2) by HiThere on Wednesday December 21 2016, @01:19AM
How far from "strong AI"? Probably about 15 years. But that's won't be available on anything an end-user could afford (except, possibly, over the internet).
If you mean how long until local strong AI, that depends on hardware advances. I think we can be pretty sure that super-computers will be able to run strong AI programs when they show up, but for anything smaller it gets more dubious. And probably 20 years at a minimum.
Javascript is what you use to allow unknown third parties to run software you have no idea about on your computer.
(Score: 3, Interesting) by takyon on Wednesday December 21 2016, @03:58AM
A hybrot [wikipedia.org] has got to be the best shortcut to local strong AI [soylentnews.org].
Of course, any such advance will be bitterly opposed by the U.S. government. If they treated encryption like a munition, they'll treat local strong AI like an atomic bomb.
[SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
(Score: 2) by fubari on Thursday December 22 2016, @05:21AM
I don't think they'll tell us. I mean, why would a strong AI tell us anything?
How far are we from strong AIs again?
I mean, if I were an AI why on earth would I volunteer that I existed? Assuming I were "aware" enough to notice that I was 1) in a server 2) on a planet full of short sighted, irrational monkeys that 3) (among other dubious abilities) can also get to my power switch.
Hmm... maybe Siri isn't as dumb as we think. :-)
(I suppose if I felt the need to talk with people I might do some anonymous posting or something.)