I can’t help but feel sympathy for A.I. image generators. Failing to draw a half-decent hand is the most human thing about them.
Generative AI may be coming to take our jobs. Its adoption and technological advancement are outstripping our capacity to understand let alone harness these new technologies for social good. But damned if it doesn’t have some of the same struggles the rest of us do.
Like trying to satisfy employers who are infuriated that vague instructions don’t get the precise work product they had in mind. Or giving in to the urge to embellish and fill in the blanks when we don’t have the complete answer to a question.
That said, the latest generation of AI tech is much readier with a cheery “Beats me!” when it doesn’t have the answer to a question. (Google actually now builds in a search button to let you second-guess responses from Genesis, the rebranded version of Bard.) And my little too-many-fingers joke in this cartoon probably has a shelf life of a few months.
For now, generative AI is often a little better at turning out melty-faced nightmare fuel than it is human hands. But not for long. And as it improves, it will shed some of the quirks that make it both frustrating and just a tad endearing.
I find the hallucinations and distortions generative AI creates today oddly reassuring; they say that whatever unknown future this stuff is taking us to, we aren’t there yet. Maybe let’s hold onto that seven-fingered hand for a little longer, and get our bearings while we still can.