Strata Lucida

I dunno. In the original image it was hard to see where some parts left off and others began. Now it's easy. From an info-conveyance perspective, the illustration carries more info than the photograph. How that illustration is created is not really important to the viewer. And to the person who has the responsibility of creating or commissioning the illustration, a good deal of not just labor, but back-and-forth communication ("Is this how you want it? No, more like this. Oh. How about this? That's getting there, but emphasize the manifold a bit more....") is saved.

Save a lot of time, and arguably get a better illustration than you would have otherwise (it sounds like you're saying that an illustrator could do a better job; I'd say that depends on a lot of factors). If this technology produced crap illustrations, I'd be inclined to agree with you, but these are extremely good.

Is this circular logic?:

> Accurately perceiving the important aspects of a car's engine is difficult: that's why we hire illustrators to do it for us.

I would say that if the technology makes what was once a difficult job easy, then we may not need to hire illustrators.

Actually the tool's output isn't the nice colored illustration we both used in our posts; it's the skeleton image in the top middle row. So the tool in its present state of development still requires a human to look at it, clean it up, add color, etc etc.

Regardless of whether the tech is fully baked, what I am offering here is a lament for the loss of a social process (the back-n-forth "more like this, no, more like that" between artist and client) and the externalization of a cognitive/creative process.

I would say that how the illustration is made *does* matter to the viewer; it's the same reason why handmade cookies taste better than machine-made. And if the illustrator isn't entirely displaced by this technology -- say, she still has to sit there and clean up the image -- then it becomes more like using cookie dough from a tube. Either way, it cheapens the activity, in both senses of that phrase: it makes her work less skilled, and therefore cheaper; and it makes her work less dignified, to boot.

At our magazine, we use a lot of illustrations that are made by tracing lines over photographs, then removing the photograph and filling in the lines with colors to produce a finished line drawing. It is incredibly time consuming and tedious work. I think that automating this work, even partially, would be a welcome development for our artists. Your nostalgia is misplaced: Like nostalgia for rugs hand-woven by small children. In some cases removing the human element is actually an improvement.

BTW, the back-and-forth with the client (in this case editors) that you're talking about happens at a different level -- more about overall look, colors, style, or specific elements-- rather than placement of this line or that line. In fact, I can see how a tech like this would allow *more* time for back and forth, partly because changes are easier to make and partly because there's more time to make them. In other words, the technology is one that augments human creativity ... potentially ... though there is always the danger, I suppose, that legions of line-drawing illustrators will be laid off once the automated computer draughtsmen are installed ....