One of my all-time favorite books is Robert Heinlein’s “Time Enough For Love,” and one of my favorite characters from that novel is Dora, the intelligent, sentient, personalized computer. If you’ve read the book, you’ll know why many of my computers around the house have some version of “Dora” in their name, as in “Desk Dora,” “Laptop Dora,” and so on. The other are named things like “Lapis Lazuli,” “Lorelei Lee,” or “Minerva.”
Heinlein wasn’t the first to think of such a thing, and the idea of a sentient computer to run a ship but also be a companion and friend as opposed to being a machine or a servant is hardly unique to Heinlein. But Dora’s always been my favorite.
We’re a long way from such a level of functionality, of course. On the other hand, we’re starting to see the first inklings of that sort of thing with programs like Siri and Cortana. Speech synthesis has been pretty good for a while. Voice recognition is getting better and is pretty functional in normal use. Image recognition is still waiting for the big breakthrough it needs, but there’s some progress. Search engines, led by Google, are excellent. But when all of these pieces are put together and the machine or program tries to function in the real world, they all (except IBM’s Watson, perhaps) get left in the dust by any bright five-year-old kid.
Thus we find that Siri (I’ll use her only as an example since I haven’t used Cortana yet) can talk to you very well and she can usually understand the words you say to her – but figuring out what you MEAN or want her to help you with is still hit and miss.
I love playing with Siri to see what she can do and what she can’t. I love finding the “easter eggs” that have been programmed in by the Apple software engineers. (Have you asked Siri, “What is zero divided by zero?”) I want it to have access to as much voice-activated, hands-free digital assistance as I can get and if that means that I have to learn how to phrase questions or give commands a certain way, so be it. I’m willing to meet the computer halfway – even while I understand the average person can’t or won’t learn a particular way to interface with the computer. (See “television remote control” for reference.)
I’ve learned to do simple, but useful things. “Siri, please set an alarm for seventeen minutes.” “Siri, please read my email to me.” “Siri, what is 45°C in Fahrenheit?” “Siri, what is the score of today’s Kings’ hockey game?” “Siri, when do the Chiefs play football again and who is their opponent?” These all generate useful answers.
But ask, “Siri, please read today’s baseball scores to me,” and you’ll get a screen full of data, which is correct, but useless if I’m driving.
I’ll often ask Siri a series of questions if I don’t get a useful answer for the first one. At first I’ll be trying to figure out how to ask it “correctly” so that Siri understands what I want and gives me a useful answer. But after a few tries, if we’re getting nowhere fast, I’ll start to let the snark out and ask stupid and occasionally insulting questions just to see how the programming handles it.
This was happening on Saturday, when I asked the following and got a hilarious (but perfectly accurate and logical) response from Siri:
The most hilarious part of all is that somewhere hidden deep inside my calendar lurks that reminder. I’m sure Siri will pick the most inopportune time (such as the middle of a job interview) to pipe up and either tell me it’s time to do that, or to inquire whether or not I have completed the task yet.
One taunts our digital “assistants” at one’s own risk.