I have a lot of gadgets. MP3 players (got an iRiver Clix to check out too!), cameras (still need a new one…), convergence products, media servers, media players, etc, etc, etc. Not a single one of them can learn a thing from me, and I don’t understand why. Before you get too skeptical on me, let me be clear: I am not talking about “learning” as in artificial intelligence (AI), I just want them to have a few behavioral modifications.
My first example is my new Harmony 880.
Works great, about 90% of the time or so. Every now and then it misses a command, or in some way gets ‘off’ with the living room. The nice thing about the product is it has a Help button which basically steps you through the sequence of remote commands necessary to perform the function you are looking for. So, if I click “Watch Moviebeam” (this is an ‘activity’ much like a macro), the Harmony turns on the Sony receiver, Syntax LCD TV, Moviebeam and my DVD player (which is wholly unnecessary, but I can’t figure out how to remove this step), switches inputs on my Sony receiver to the correct audio input, switches inputs on my Gefen 4×1 HDMI switch, and switches inputs on the TV. Awesome.
But when I first started, it could never get the Gefen to switch right. So I clicked Help, and followed the on-screen ‘yes/no’ options until it worked. Eventually I went back to the Harmony setup application on my PC and found I could ‘slow down’ the commands to make it more reliable (which did in fact work like a charm).
I think it would be very easy for Logitech to add a bit of intelligence to the Harmony. Maybe after doing the same thing 3 times the remote could say “please dock to your PC for an update”, at which it uploads the info to the PC, and the PC application can have enough smarts to say back to me: “there is a problem with your activity, let’s try to fix it.” It doesn’t have to be smart enough to actually fix it independently (although that’d be nice too), but it should be smart enough to track little things like this. Extra points to tivoboy who seems to agree.
My second example is my Garmin Nuvi 350 GPS unit.
Again, a device I absolutely love (full review still forthcoming, sorry, too busy working to pay the SF rent) and use all the time (as does my wife!). The Nuvi is absolutely great at getting us around town and the Bay Area, helped me go camping, helps my wife find her delivery locations, we love it long time (and we certainly do not leave it in the car overnight). I really like setting ‘favorite’ locations (pictured to the right), for work, for home, it’s quite handy. In fact, its versatility at finding locations is quite impressive.
What I don’t understand about it is the inability to learn just a few things. Remember, the Nuvi 350 can play MP3s, audio books, show pictures, find nearby ATMs (and more importantly, Dairy Queens) and even pump gas and check the oil pressure for you. But no matter what I do, it refuses to remember the fact that I live in San Francisco. Every single time I enter a city name, I have to type in “San F” and then select from a list.
How about, after the 5th time I make that choice, it simply auto-saves it, and then give me an option to ‘Change City’ next time? It even has the concept of a “home address” and knows I live in SF (can’t call it San Fran or Frisco, they hate that here… I don’t know who ‘they’ is, but I know I will get whacked if I use those terms). I’ve got two more ‘freebies’ for Garmin as well:
- If I click ‘detour’ give me the option to keep that detour. Maybe it’s a closed road, or just a really bad route. Better yet, if I deviate from a route multiple times, ask me if that’s a better way to go! The manuals state very clearly that the Nuvi will not guarantee the best way based on city shortcuts and everything. I totally understand and appreciate how much work that must be. But why won’t it pay attention to my driving and catch on after a while?
- When I turn on Nuvi, I always fall into one of two behaviors: I either wait 10-30 seconds then click View Map (this implies I am already driving), or I almost instantly click Where To. How about, after 20 seconds of inactivity, it automatically goes into View Map?
Now I know learning isn’t easy. I know AI is ridiculously hard (although improving every year). But even Windows has some basic learning features, and everybody likes to rant on how hard it is to use. Office automatically adds contacts. Word adds, well, words.
I think the gadget guys need to take a few cues and have their devices do so as well.