It's interesting to me, the model does not know what a house is like, it knows what <i>pictures of houses</i> are like. So it does a good job of making pictures that look like pictures of houses. But if you look closely, a lot of the details are really weird, unbuildable, or just non-sensical.<p>All of the image-gen models have this problem - look at the hands and faces in the generated images of people and there are often bizarre deformations.<p>It's fascinating because it's the opposite of how children learn to draw. They tend to think about the pieces that make a thing and then try to put all the pieces on paper, and they end up making a drawing that (for instance) looks nothing like a person but has two eyes, a nose, a mouth, etc. in roughly the right relation to each other. (They rarely draw ears though!) The child is thinking about "what makes a face a face" and then trying to represent it.
The ML model is sort of distributing pixels in a probabilistic way that comes up with something very similar to the pixels in a sample image in its training set, superficially much better than a kids drawing and yet in some ways much worse upon close inspection.
I don't know if it's just me, but it seems to only generate luxury homes in a contemporary minimalist / modernist architectural style. Not once did it generate anything resembling a Tudor mansion or Victorian cottage or Midwestern rancher or split level.<p>Nothing really wrong with that and it's neat anyways, but seem to represent its own sort of strange bias in a sense.
Neat, but feels like it was trained or prompted for luxury homes. I clicked threw a couple dozen images but everything had that style of multi-million dollar mansion/condo with a pool and only glass for walls. Nothing even remotely suburban or rural. Curious if this is due to the prompts (assuming this is SD sourced) or due to the training of the AI model.
I feel like we are at the tipping point where the ability to generate scenes/faces/houses/art that can reliably fool human senses becomes commonplace.
Noob questions:
Are these images that cheap to generate (once the model is trained) that we can get a new one fast on each browser refresh?<p>Are all the _this x doesn't exist_ generators based on similar scripts, only the training datasets differ, or are they somewhat tuned for the specific domain?
I'm waiting for the crossover with McMansion Hell where somebody uses similar image generation to generate the exteriors and interiors of McMansions.