Imagine, if you will, an earnest old man trying to live a young life in the Foetal Age of AI, and failing.
Oh, sure, the old man and the keyboard had some fun ideas. Create AI images. Animate them with AI video. Use them both to promote a new online endeavor called, say — PrairieVoice.com — and utterly failing in some attempts that only served to remind the old man that, yeah, even after 35 years of earnest trying… computers still suck!
First, take a look at this AI image. It’s a whole mess, right? Can’t read it, really. The AI Art Bot was instructed to “create a Prairie Voice logo” and after hundreds and hundreds of attempts, this was one of the irreverent results.
But the insanity doesn’t stop there… here’s what’s really going on…
The computer youngsters will simply tell you create the text image in something like Photoshop. Or, they’ll argue, write a better prompt for the AI Bot. Yes, that’s all true, but what fun is that? I like sparse prompts because it forces — frees, actually — the AI bot to try to think beyond its own box and create something, ANYTHING, to answer the prompt; and that’s where the oddity magic happens in the vulnerable attempt to return something AI that is not fully understood by the AI.
So, I sometimes get a mess like that image. You can sort of make out “Prairie” but not really. I kept the image just so I could share that AI process of creation with you.
Okay, so we have that icky image, in my day, we maybe called it “funky” — so let’s take the next illogical step and see what another AI Bot, the Veo video bot — might do to animate this curious image. Sure, I’ll give it some perfunctory PR words for the VO, and maybe then the video bot might just turn this weird text into something animated and readable?
Barf!
I admitI laughed at that existential animated video version of a failed Prairie Voice logo; but what a perfect AI Bot response to something so fundamentally incomprehensible?
So, instead of playing along, the video bot gave me something just as silly and ridiculous as the image it was instructed to animate: A young woman with only a head and no body, appearing in the middle of the text mess to provide us with the slug line of this new Prairie Voice endeavor!
It’s either genius or stupid — and I’m hoping for both!
Getting a video bot to animate any image is a challenge. Yesterday on social media, I shared a video bot attempt at rending an animated version of this original image:
And here was the result, along with my canny social media blurb:
It took FIVE tries to get Veo to pronounce the "Dot" in "Prairie Voice Dot Com." Four times it said "Dote" -- I have no idea why a tech company would ever program "Dot" to be "Dote" in the first four attempts!!
The sad part about all of this is that Veo — even on Gemini Ultra! — only generates FIVE videos a day; so do you live dangerously? Do you keep regenerating (where the new video replace the old one), or do you “feel lucky, punk?”
The punk usually wins. I like to keep trying until something right happens. Expected the expected, but use the insane.
So there you go, and here we stand. Alarmed and misunderstanding. Attempting to comprehend the virtual world surrounding us.
Do we believe our own eyes, our own ears, or our own fingers furiously typing prompts that may, or may not, return us to the ridiculousness of reality that has become the substance our unpredictable lives?