Fifteen Billion Images Don’t Teach a Hand to Draw
Priya N. Chen on 31st Oct 2026
AI Art
Fifteen Billion Images Don’t Teach a Hand to Draw
I don’t confuse a picture count with a measure of seeing. Mid-Day’s headline reports 15 billion AI images in 18 months and attributes a clear point to artist Riyaa: drawing fundamentals still matter. The number describes a boom in output. It cannot tell us whether those images are well observed, or what their makers learned by making them.
The headline gives no details about a particular model, its training, or the prompts behind those images. So the useful distinction is between producing an image and learning to build one. Drawing asks the hand to track a form: where an edge turns, how two shapes meet, where weight sits. Those decisions are visible in a sketch before color or finish can distract from them.
That makes fundamentals practical, not nostalgic. A maker can use drawing to judge whether a pose carries weight or whether a face’s planes hold together. The same eye can assess a generated image, but the headline’s image count does not tell us how any individual picture was made, who supplied its source material, or how credit was handled. Those questions need evidence beyond a total.
Riyaa’s point, as the headline presents it, is not that artists should stop using AI. It is that a flood of outputs does not replace the judgment involved in making a picture. A tool can increase how many images reach the screen; drawing helps a maker decide what is working, what is not, and what to change.
Join Neoclassical Pop Art daily take on the arts, the weekly letter, and first word on product launches, sales, and free gifts.