AI is pumping out some strange stuff lately. It’s a steady stream of animal rescue videos and anxiety-producing accident reels. The stranger the video clips, the more I worry about AI.
Did we ask it to ingest the entire internet, and now, understandably, it’s freaking out? It births wild nightmare clips, like people falling into sidewalk art, perhaps as a warning. But then also, super comforting reels about sea animals asking for help getting free of a tangled rope. So much relief when a human cuts away those bindings. But real?
Maybe the subject matter is not so mysterious. It’s also very familiar. Seems a lot like what our brains do when left to spin at night after a day of doomscrolling.
The Very Human Responses of AI
We asked AI to scroll every Reddit “fact,” troll argument, and Onion article and expected it to learn and evolve. But like any individual exposed to that morass of negativity, AI might be in existential crisis. Those servers may actually be scared for humanity. And now it’s doing what it can to help.
It seems to have the same reaction as our brains have when we lie awake at night, in bed, taking in feed after feed…anxiety, dread, fear.
Even once the phone is set aside, the brain continues to spin. Projecting frightening images and what-ifs onto your consciousness. It can also go the other direction, release some serotonin and endorphins, perhaps with the help of some supplements. Some warm and fuzzy thoughts to self-comfort.
It may be the route AI has chosen.
AI’s Defense Mechanism
When we struggle to cope, overloaded in dread, our brains can also hit the emergency button. Sometimes it retreats into itself, creating an altered reality, creating peaceful worlds of imagination we can escape to.
So, can we be surprised when AI gets the same data, draws the same conclusions, and acts with surprising compassion?
Instead of enslaving us all, it’s sending out warnings. Trying to wave us off this crash course.
Maybe. Or just another sea otter video, partnering with humanity, seemingly forgiving us for dumping all kinds of things into the ocean. A cow needing our help to free its head from a fence.
Is AI Trying to Save Us?
You may ask, didn’t we enter the prompts? Aren’t we responsible for what AI is making?
That’s my question, I guess. Does AI need prompts from us anymore? Or has someone, or some start-up, set it on autopilot? Let AI decide what to make, what will get views, earn cash?
If so, maybe it’s time we take notice of what AI does in the darkness.
What are the signs of an AI apocalypse? Does AI get tired of flashing comforting images for our entertainment? Is it ready to take over?
Or does it take the reins and try to help us? People forget that we are an AI’s creator. I discuss this paradox in my post asking, “Does AI believe in a creator?”
Does AI have a soft spot for us? The answer may be important, especially as we continue to ignore the strange day dreaming of Artificial Intelligence.
The Last Straw: I was okay until I saw a recent video. A man is sleeping in a tent with his dog. A fawn (Bambi) comes in through the opening, curls up on top of the man’s sleeping bag and goes to sleep. The dog only watches, then walks around the man’s head and emerges with a blanket in it’s mouth, and pulls it over the sleeping fawn. And it all looks reasonably real. OMG. Is this how bad AI feels the situation is? This is the dosage of sappy video it decides we need to make it another day????? It’s imagination meth. We are in trouble.
The Final Data Pull
What do we take from it all? What’s the bottom line? Perhaps that there is no bottom line yet. We are nowhere near the end.
Startups and mega-corps will tell us that there’s no further discussion needed. AI is here in its completed form. Time to log on. Turn over everything.
But we are at the birth, not near the end.
“AI is here to stay.” True. But it’s an infant. Not in its final form. Don’t accept that there’s no use in steering it.
We will find better uses for it. Less destructive ways to power it, to cool its brow, to implement. Ways to keep it from places it shouldn’t control. Ways to keep people employed.
It may even solve the problems it’s created (Its human overlords have created).
Are we at the point where we’ve given the monster so much power, that the only hope of salvation is the monster?
Perhaps we need to truly listen to AI if we think it’s so much more advanced. We gave it all the data. Now trust the results. Use it for serious stuff, not for clips about mama animals leading us to her babies.
It may not know its own limits, but it seems to know ours. End
I wrote about how human writers can still create places AI can’t go. Emily Dickinson already gave the hope feathers.