Behold, your toddler’s next Power System client device will be a Watson-connected toy dinosaur. But am I the only one who finds the idea of letting a global megacorp’s A.I. communicate directly with our children slightly creepy?
I ask because among this week’s media coverage of start-up toy firm Elemental Path’s plans to do just that, there has been nary a dissenting voice. Instead, the press found the idea of its prototype CogniToys kind of cute.
On the face of it, the CogniToy is a pretty clever idea. Cheap too. There’s not much going on inside its green shell other than a speaker, a mic, a battery and a direct connection to a big cluster of Power Systems in the cloud.
Those Power Systems run the advanced analytics and voice recognition software that make up Watson. Proud CogniToy owners will be able to chat away to world’s most advanced commercially available A.I.. And Watson will chat back. And listen. And learn.
I can’t help thinking that if this were a plot in Doctor Who/Person Of Interest/Fringe (delete where applicable), The Doctor/Finch/Walter would be doing their damndest to blow up the toy factory concerned and/or change the time stream so the venture never gets off the ground in the first place. In fact, at least two of these TV shows have featured uncannily similar scenarios.
Perhaps some of this unease is the paranoia of being a parent. It is interesting to note that CogniToys’ Manhattan-based inventors are themselves childless. If they weren’t, they would know that some of the best, most difficult, questions ever asked come from your kids. These can range from the existential to the philosophical to the disarmingly practical.
Trying to at least attempt to answer them is part of what makes us humans. Abnegating such tasks to the seemingly innocuous electronic gewgaws that litter our lives seems, at best, unfulfilling. At a time when Samsung has just admitted that its TVs can act as constantly eavesdropping Orwellian telescreens, it also seems slightly menacing.
Elemental Path reckons that it has built in functions to its prototype toys that will tell a child to go and ask a grown up if the questioning gets inappropriate. It calls such enquiries “mommy questions”. But, again, aren’t such moral navigations really best left to individual parents or carers? It’s all part of the fun and, anyway, who decides what’s appropriate?
While I can accept that Watson may well be able to shoo away a kid from obvious subjects, I’m not so sure I would want it to be informing my child about very many others.
IBM’s Cognitive Computing Engine is, after all, the construct of a 100 year-old non-unionised American corporation. Its world view is bound to differ somewhat from that of a liberal European like myself. Especially one with Stephen Hawking’s recent warnings ringing in his ears about how artificial intelligence could end mankind.