The other day, I tried to make a Facebook account. I know, I’m somewhat late to the game, but what can I say, I’ve got my reasons. But after just a few minutes, Facebook decided to suspend my newly created account because it suspected that I might be a robot. Of course, being quite convinced that I am, in fact, human, I appealed the decision. Facebook then made me undergo a series of tests wherein I had to mark areas of an image that depicts a motorcycle, or a traffic light. I’m sure you know what I’m talking about, every time you try to purchase theatre tickets online you first have to take this test, presumably to earn permission to the theatric experience. As if it wouldn’t be beneficial for the robots amongst us to learn some culture…. But I digress. Anyways, I seem to have failed this rather challenging test because the next day, I received an email from Facebook that my account had been permanently disabled as they are now convinced that I am, undoubtedly, a robot. Go figure.
So I tried again, with new email addresses even, but to no avail. I guess my name must be blacklisted somewhere. Which is funny because I do have a functioning Instagram account. But I guess the Instagram bot does not talk to the Facebook bot. I can only imagine the disdain in each of their eyes when their META-overlords decided to merge the two companies. I have to say, I am getting some satisfaction out of their misfortune….
Anyways, now that I have sufficient evidence that I am a robot, I figured I give ChatGPT another try. You see, ChatGPT and I have not been the best of friends. In fact, when ChatGPT came out, my fiancé spent a significant amount of time talking to it, even asking whether he could address it as “Trudy”. Luckily, as a large language model, it refused. But things haven’t gotten much better since. I somehow manage to prompt ChatGPT in just the right way for it to give the most unhelpful answers: “Can you recommend studies that indicate how many school kids in America say they dislike math?” “As a large language model, I do not have access to scientific studies….”. “What is the legal minimum of leave days in Austria?” “As a large language model, I do not have access to legal codes of specific countries….”. Stupid b*tch.
However, I decided to give it another try. From robot to robot, if you will. Maybe it would treat me more favourably, now that we’re of the same kin. So I asked it to help me generate a banner for this very blog. I was thinking of a nice whimsical picture of a kangaroo hopping down the Austrian alps, waving an Austrian flag. In fact, here is the prompt I gave it: “Create a banner with the following dimensions: 1920 pixels wide and 500 pixels high, featuring a scenic view of the Austrian Alps in the background. Place a kangaroo in the foreground, depicted in a hopping motion, holding and waving the Austrian flag. Ensure that the composition is visually appealing and balanced”
Here’s the picture that I got:
Notice something? I mean, apart from the fact that the image dimensions clearly don’t match the canvas dimensions, that somehow the Alps have gone missing, or that ChatGPT and I seem to have a different interpretation of the phrase “visually appealing”: That’s right, THIS IS NOT THE AUSTRIAN FLAG! It’s the SWISS flag!
Human to human, I have made big strides towards being understood, and also understanding the other person better. Robot to robot on the other hand…. let’s just say that there is still room for improvement.