If you’re like me – you do your fair share of social media surfing. I don’t know how many times a day I come across some outrageous quote (usually tied to an inflammatory or funny picture) that basically screams “click bait.” Those are easy to spot, and I can easily make a choice on whether I want to play along or not.
Then – there are the other types of posts – where things *sound* like they should be (or could be) correct – but the source is some site I’ve never heard of. They usually have passionate comments both pro/con on the matter – and inevitably – I’ll have an opinion one way or the other.
Usually, I keep to my motto – “Opinions are like a**holes – everyone has one” – and just keep scrolling.
Usually, a simple search will tell you whether the story (or topic, or quote, or action) is a real thing just by the number of other posts that turn up. I always scan the site URLs in the search results looking for “reuters” or “apnews” or other news sites that I have found to be fairly correct in the past.
If I do happen to find a few different sites that confirm that the story (or topic, or quote, or action) has actually happened, then I can always choose to re-engage and regale the world with the depths of my superior intellect and scintillating wit…
So, I sort of fell down the Internet rabbit hole while researching how these things get started – and how they spread.
The bottom line is: people just make crap up and post it.
So, I wanted to see if I could just make up my own realistic-sounding BS on a number of different topics and see what would happen. I quickly realized that I had a few problems:
- I’m lazy
- I have a limited imagination
- I really despise bullsh*t
So, in my geek-quest to at least get a couple of good quotes – I came across a website that uses AI (Artificial Intelligence) to auto-generate a bunch of crap that sounds like it *could* be real.
Here’s some examples (the text I wrote is in bold everything else was 100% computer generated). First – I went for the low-hanging fruit – something political:

Not too bad! It even came up with fake quotes from fake lawmakers. Next, I changed subjects to technology:

It’s a little “around-the-block” but perfectly plausible. Next, I went for a little fiction and drama:

The great American novel it ain’t – but it’s an interesting snippet. I wondered if I kept the same quote if the AI would come up with something similar or totally different.
The answer? Totally different:

Finally, I tried something a little more “open ended” – and was not disappointed:

The generator even referenced a fake Twitter handle and fake URLs to fake tweets to “prove” that I said something I never did. So, of course, I had to “poke the bear” and try my luck again. This time I was just “part of the (political) story”:

If you know me, you know that I couldn’t just leave well enough alone… so I tried the same phrase one more time… and… well… the wheels came off the bus:

Moral of the story – you can’t always believe what you read.
You can try the “Talk to Transformer” for yourself here: https://talktotransformer.com/