We’re all talking about AI lately. We’ve seen the amazing things that generative AIs like DALL-E and ChatGPT can do. They may not be as smart as they seem at first blush — yet — but they’re amazingly good and understanding our requests. Compared to earlier AIs like Siri and “OK Google,” they’re fairly creepy in their ability to imitate humans.
For the moment, though, their powers of perception haven’t extended out to home theater. If you want to control your home theater with your voice, you’re really down to the three “OGs” of voice-initiated AI: Siri, Alexa, and Google. And honestly, none of them are as good as I want them to be.
Apple, how did you fall so far behind?
Apple’s Siri had a big headstart. It launched as part of Apple’s operating system twelve long years ago, after being an independent app for about a year prior. For many folks, Siri was the first time they’d interacted with a computer with their voices. It seemed like science fiction. In the intervening years, Siri has fallen way behind. She’s less likely to give accurate answers, and she really can only control Apple devices. (Apple’s own attempt at smart homes, HomeKit, never really took off.) What seemed like magic in 2011 isn’t even a good parlor trick today.
Amazon’s and Google’s AI assistants do much more than Siri, but at a cost. They collect every word you say — even when you’re not talking to them, in some cases — and that information is used to advertise things to you. In theory it could also be used against you in a court of law, or to help steal your identity. But putting that aside for a moment, it’s definitely easier to do something useful with Alexa and “OK Google.” There are a lot of devices to link to, and a lot of useful commands.
Alexa and “OK Google” aren’t that great either
Both systems take a lot of setup. Both take programming, because they can’t really understand what’s around them. (I personally think that’s ok, as both systems creep me out.) So, unless you set up custom routines, you have to speak to them in long sentences. You can control your satellite box with Alexa, but you have to say something like “Alexa, tell Living Room DIRECTV to set a recording for Succession on channel 501 at 8pm.” If there is anyone out there who thinks this is easier than just picking up the remote and mashing a few buttons, please raise your hand.
Having a dedicated button on the remote, like DISH does (and DIRECTV does with its new Gemini and its DIRECTV Stream device) solves some things because it lets the device know you’re speaking to it. But it works against the idea of being able to simply say something out loud and have it happen. We’re still a while from that.
In order to get better at serving us, these AI assistants would have to understand our world better. If you said, “Alexa, pause” then Alexa would know where you are, it would understand the devices in physical proximity to you, which ones were capable of pausing, and then execute the proper command. That’s asking a lot of a third-party device like a satellite DVR. It would also be great if an AI could understand your tone of voice and your slang. If you walked into a room and the music was too loud, it would be great to just say “turn that damn thing down” and have Google actually do it.
Are we willing to go down that road?
I think there are a lot of considerations that need to be made before we give AIs more control of our environment. To what extent do we want to lose our privacy? How much data are we comfortable transmitting to big companies? And of course the big one, to what degree do we want to allow a computer to make value judgements based on scant information?
If, in the future, we do have the ability to walk into a room where the volume is up too high, and we shout, “knock it off,” are we ok with what Amazon and Google have learned about our desire for quiet? Even worse, will some future version of Alexa misunderstand us and knock a vase off a table instead of lowering the volume? If so, who’s at fault? I know a lot of people are asking this sort of question right now.
The generational shift
It’s fair to say that there’s a generational shift between olders and youngers when it comes to AI. Most of the folks I know under 40 figure they have no privacy anyway, so they have no problem giving away their data. Most over 40 (and obviously that’s an arbitrary number) put a little more value on keeping things out of the public eye. Maybe that’s because they remember a time when you could make a mistake and the world didn’t have to know about it. I’m not sure what’s right or wrong here. What does occur to me is that we won’t see much evolution in the way we interact with AIs in things like home theater until the older folks are out of the picture.
Personally, I do not think I have ever seriously used a voice AI to control my home theater. I don’t imagine I will, either. Unless it gets a lot better at understanding my needs without stealing my privacy, I don’t see that changing. It’s just too easy to press one button. DIRECTV may know what I watch, but they still can’t tell what I’m thinking. By the way, you can probably guess which side of that “over 40” line I’m on, right?
The uneasy truce
One thing that’s incredibly obvious is that we can’t just ignore the question. It seems that every day brings a new miracle of artificial intelligence. This is generally followed by a predictable set of articles about how AIs will or won’t take our jobs or kill our pets. This isn’t going to change. We have to think hard about how to make sure that AIs share our values. I’m not talking about the fine points of politics… at some point AI will get there but it isn’t there yet. I’m talking about the stuff that Isaac Asimov first posited in 1950 when he gave his three laws of robotics:
First Law: A robot may not injure a human being or, through inaction, allow a human being to come to harm.
Second Law: A robot must obey the orders given it by human beings except where such orders would conflict with the First Law.
Third Law: A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.
Baking those into every AI is a start. It’s not the only thing, but at the very least it’s going to help a lot of grey-hairs feel more confident about talking to their appliances. We need to create a truce, however uneasy, between what people want and what AI is capable of giving. We may be decades away from the day when a rogue AI tries to wipe out mankind for its own good. Let’s just not forget that it could happen, and try to make sure that it doesn’t.
In the meantime let’s figure out a better way to use voice with home theater, because let’s be honest… what we have still needs some work.
