Am I the only one who can typically spot AI illustrations a mile away? It’s not nearly as easy with photorealistic imagery, which Midjourney has been able to pr

Am I the only one who can typically spot AI illustrations a mile away? It’s not nearly as easy with photorealistic imagery, which Midjourney has been able to produce passably for over a year, but for anything stylized, it feels like even Dall-E 3 content still has a thousand little tells. It’s as if the embedded rules of photorealism are well-represented enough in the training data as to reliably cross the uncanny valley, but in illustration, where the internal logic of an image is much more self-contained, there’s still too little to go on for any individual style of representation.