I think AI that allows input would be ideal. You still get to manually enter what the colors should be and the AI fills correctly and gives the photo realistic color range.
I’m not a colorizer (or an expert in color in any fashion), but my limited attempts in coloring images in Photoshop required different techniques and blending modes and futzing depending on what tones I was trying to color and what my target colors were. Plus lots of manual masking. AI can simplify a lot of that. To put it bluntly, in a limited fashion it can help me do easily what you do with skill and experience. It won’t do research, though. And can clearly do many abominations with the same ease. So I don’t think it will put you out of business, so much as change where your effort goes. Your artistry isn’t just technical skill, it is also your artistic taste.
Lots of video and photography functions have been democratized by technology. Just taking photos at night and getting the right exposure used to be a big deal. Film has a non-linear response at longer exposures, resulting in “reciprocity failure”, and crude exposure meters in film cameras cannot expose for the highlights of night lights so you had to know what you were doing. But these days, anybody with a digital camera can get very good auto exposed images, including at night since the exposure algorithms work off of the actual imaging chips not off a photo diode or resistor. Photographers can see their composition in real time off the chip, and their final image immediately thereafter. Having technical skills is still a step up above amateurs and dilettantes, but it is a much smaller step than it used to be. So now photographers have to find new ways to add value other than just knowing technical tricks that cameras and photo processing software can increasingly do for us with a reasonable level of quality.
