[waddlesworth]

I definitely could be absolutely wrong about this, but I don't think computational photography will be a huge way forward for these kinds of cameras.

Smartphones have fully taken over the light/compact segment, they're better at it, you always have it with you and they've gotten good enough that for anything 'happy snap', a smartphone will do the job fine.

The professional/high end hobbyist segment is still going strong, and I think is growing steadily. The appeal, for me at least, is a lot about the creative control. I'm the one creating the image, not the device I'm using. There are computational tools I can use, but I'd want to do anything like that after the fact, and Photoshop has added some tools pushing this forward, but I do think there's a lot of room for a more ML focused editing experience.

There is a certain amount of push for computational features that help with taking a photo, like fast autofocus that can see a face and focus on the eyes.

When it comes to actually generating or modifying the image, I think this is a line that most photographers don't want to cross.