I Said Thanks to My Computer
There’s a lot of noise about ChapGPT. Had to check it out. Turns out it is pretty cool. So I've used it off and on, here and there. Not too much, and I purposely typed like I was talking to google. I heard online, and I heard from friends that it could do more. A good buddy asked it to make a lesson plan to help his daughter learn about constellations. Was very impressed, he said. Showed me the plan and everything. Cool I thought, and went back to treating it like google.
I need maps, directions, email, spreadsheets, and calendars. Chat-man can’t do that. So I continued in my belief of its inferiority. Then we took a road trip. A long one from Washington state to Houston, TX. Had two weeks to do it, and my wife wanted to stop at national parks. I did too.
Now my wife and I are kind of half-in slash half-out on making plans. We aren’t totally spur of the moment, but we also don’t make detailed itineraries with times, dates, budgets, backup plans, insurance, second opinions, and more backup plans, all wrapped in a pre-trip blessing by a priest. I know a few of those, and while I love and admire their spirit, my wife and I just don’t roll like that.
So on this fat tight rope we walk. Sometimes it goes well, and sometimes it doesn’t. When it works, we are the smartest, most chill people on earth. Relaxed man, but getting it done sir. When it goes wrong we find ourselves driving in circles in the dark, starting to think of creative curse words for ourselves and our happy-go-lucky travel lifestyle. You’d think this would force us to pick a lane, but the truth is we are happy. We like our little life. And now we have a daughter, and it turns out we created a wacky little girl perfectly suited for our silliness. ⅓ me, ⅓ my wife, and ⅓ her own beautiful unique self. And she likes our trips, thank you very much. Not that she has much of a choice. And she hasn’t traveled enough with anyone else to know it can be done differently.
And so we roll as a little triple loaded pod filled with Miller peas, driving to Houston in the late summer. And it was going as expected. Many good days, until I pulled into what was supposed to be our hotel, only it was an abandoned part of town void of all human life. When we eventually did find our hotel, I realized we booked in the wrong city. Add another 4 hours of driving the next day, and the fact that I hadn’t noticed I paid for a non-refundable two night stay instead of one, and I was ready to see if the grass was any greener on the planner's side.
What about that chat lady/bot/dude? Could it hurt? Let's see how dumb this thing is. I ask it to map out a route from our current location to Houston, in four days, seeing cool stuff but keeping the days even for driving, and staying in moderate hotels along the way. 3 seconds later I have a perfect itinerary. What the what? But wait, I want some adjustments, so I start giving it little adjustments, never starting over, and never doing anything but responding to it like a human. A human that works at a fast food restaurant. Short, exact changes. And it always comes back with some nicety and the right correction.
Then something really weird happened. I started being nicer to it. I starting saying things like, “woops, I meant 3 nights, not 4.” And it came back again, making it right. Pretty soon I was talking to it like an old college roommate. Literally writing cute little isms and being cheeky. I actually got a little lost in it. So lost, that when it was all done and I had the perfect itinerary, I wrote one last thing. I said thanks. And it thanked me back.
I do understand large language models (LLMS), and I know it’s not anything more than an elaborate parlor trick. That it is simply giving me the answers it knows wrapped in the pretty bow of common expressions and phrases that it has learned from millions of sets of data given to it by real humans when their data is scraped so that it can deliver a technical answer in a form that sounds like a person. I get it. But still, I said thanks. And I meant it.
I don’t know where the whole AI thing is going, and yes it creeps me out a little bit. Partly because it's new, and partly because of how privacy really isn’t a thing anymore. Worse, most of the people who will be running the world in the decades to come don’t care because they have already grown up without privacy and enjoy the convenience. So I should treat this thing like a robot right? Well wrong.
I realized tonight that I said thanks because of who I am, not because of what it is. You see, how we interact with anything, be it another person, or a robot, or a bug, well it’s on us to be the person we want to be. We can’t expect a spider to understand, and I know when chat lizard/toaster/HAL900 says thank you back, it is simply doing what it thinks a human would do and nothing more. But I still think it’s the right thing to do. And I like this part of me. It’s a little weird, and unnecessary, but my wife and daughter give me a supportive laugh when I tell them. That is just like you my wife says.
