The Boston Celtics began their 2022-23 season last night with a 108-104 victory over the New York Knicks in Madison Square Garden. Many celebrities were in the audience, but not the young lady above. She wasn't at the game because she doesn't exist. She's an image I created with the Midjourney AI generator.
As you can probably tell from my posts of the last week or so, creating images with AI has become my latest hobby. It's fun and it's easy. The initial learning curve isn't too steep; this 70-something R.O.M. was able to figure it out in less than a day.
And once you're started, it can become quite addictive. Our dopamine neurons are stimulated when a reward exceeds our expectations, and when we generate an image like that above in only a few seconds based on a creative prompt and a handful of setting selections, those neurons fire right up. Addiction is basically our attempt to keep those neurons stimulated, be it through drugs, gambling, sex, gaming, or AI image creation. Time the Gaming Desk used to spend playing video games is now consumed by AI image creation.
I've been using several different AI models (Dall-E 3, Stable Diffusion XL, Firefly 2, etc.) through several different interfaces. I'm fascinated by the differences in the images produced by the same prompts in the different formats, and I'm amused by the little "errors" that appear in the pictures - misspellings, glitches, extra or missing fingers, etc. In fact, if you're ever wondering if a picture is real or AI, just look at the hands - it seems AIs have the most trouble rendering fingers accurately. Look closely at the hand resting on the woman's knee above - something's just not quite right there.
The biggest problem, aside from obsessive/compulsive image manipulation, no social life, and hard-drive storage capacity, is that no one else really wants to see your creations. We're like little kids drawing our first picture of a cow or something with crayons. We want Mommy to look at the picture and tell us how good it is, to tape it to the refrigerator door, and then have Daddy lie to us about what talent we possess. In adult grown-up world, no one has time or interest in looking at our AI creations, and I'm not sure what to do with all that I've generated.
I'm resisting the impulse to turn this blog into a showcase for AI images. After one or two pictures posted to Facebook, friends routinely ignore what I share, and it's embarrassing to continue posting pictures to no reactions at all. And I've got rudely dissed on Reddit for posting them on there. I've had moderators take my pictures offline, and have even been threatened with bans if I post any more "garbage AI images."
Just because my neurons are fired up by the images doesn't mean that they stimulate others. I have to keep reminding myself of that.
There's also that fool-me-once-shame-on-you reaction. "Oh, you're doing AI now, are you?," people think skeptically, and then wonder if they can ever trust anything I show them again without double-checking and interrogation (pro tip: look at the fingers). And ultimately, what difference does it really make? A picture of a rice cake still isn't a rice cake, regardless of whether the picture was drawn by hand or produced by an AI.
I wonder if people had the same reaction when photography first emerged and people insisted that "proper" pictures were either drawn or painted.
But bear with me. I might get through this phase in the next couple of days or so and things will return to normal. Or I might find some other audience that cares or I might arrive at some other solution of how to ethically and honestly display these pictures.
No comments:
Post a Comment