A recent article in WIRED magazine suggested that whatever we post online — writings, photos, videos, phone number, shoe size — is right now (as we speak) being used by various generative AI computer programs to build their databases.
Without our knowledge.
Except that we now know about it, if we have purchased a subscription to WIRED magazine.
But which writings and photos and shoe sizes are being used? And for what purpose? We don’t know. And we will never know.
So I did a little test. I asked ChatGPT to write a brief bio about Colorado journalist Louis Cannon. It wrote:
As of my last knowledge update in January 2022, I don’t have specific information about a Colorado journalist named Louis Cannon. It’s possible that he may be a relatively new or lesser-known figure in the journalism field, or there might be limited publicly available information about him.
As you can imagine, I found this response to be both depressing, and comforting.
Depressing, because the world’s most popular AI tool has never heard of me.
Comforting, because the world’s most popular AI tool has never heard of me.
Also, depressing, because — according to this response — the world’s most popular AI tool hasn’t had a “knowledge update” since January 2022. Like, two years ago. ChatGPT still thinks we’re paying $2.49 per gallon. It still thinks COVID is a thing.
During those two years, I have posted literally hundreds of humor articles on the Daily Post.. And some of them were halfway funny. But apparently, ChatGPT missed them all.
Is OpenAI trying to create a Generative Pre-trained Transformer with no sense of humor?
Admittedly, I don’t always keep up with the news like I should. I still don’t know who the Speaker of the House is. Did they finally pick someone?
Apparently, ChatGPT doesn’t know either. Or maybe a “knowledge update” means a “knowledge update about unimportant facts.” Which there are a lot of, in this world. And some of them aren’t even facts. Some are ‘artificial facts”. Or maybe, “semi-facts.”
That’s a term I just invented. “Semi-facts.”
How long will it take before ChatGPT finds out that, right here in the Daily Post, I invented the term “semi-facts” on January 3, 2024? Two years? Maybe never?
I’ve written several humor pieces about ChatGPT, and you would think it would have noticed me, if for no other reason.
I imagine most people will be perfectly content if ChatGPT never learns of their existence. Not everyone has an ego as big as mine. For most people, it’s enough to be known and recognized by a few of the classmates they connected with at their last high school reunion, and who are now friends on social media. ChatGPT does not attend high school reunions, as far as I can tell.
This whole problem naturally got me thinking about the internet in general, and wondering if anyone knows about me. Like Google, for example. Turns out, when I did a Google search for “journalist Louis Cannon”, the computer thought I was searching for American journalist Lou Cannon, who was state bureau chief for the San Jose Mercury News in the late 1960s, and later senior White House correspondent of The Washington Post during the presidency of Ronald Reagan.
Okay, I can understand I’m probably not as famous as Lou Cannon. As a journalist. But to be left completely out of the next 100 results? Google even listed boxer Joe Louis in those results, and I don’t think he was ever a journalist.
In desperation, I searched for “humorist Louis Cannon” and low and behold, Google thinks I’m a humorist, but not a journalist.
Be thankful for small favors, I always say.