One thing you definitely won’t see are things you haven’t seen before — that, however, may be about to change.
The new move to art by cognitive computing/ IBM’s Watson is interesting, though. Most artists will tell you that any significant work is usually very hard work. It’s manual labor, it’s chipping away at a blank piece of marble or blank piece of canvas, or some other medium. Unless you’re really in the zone and can fly along at speed, it’s tough.
The new theory is that cognitive computing is going to be a major force in future art. Cognitive computing is all about machines learning and even rewriting their own code to improve themselves. Art is very much a learning experience, so that’s an intriguing, if not necessarily very convincing, argument.
The question is what will a machine learn? How to execute lines, colors, shapes…? Will it learn composition, or simply “create”? The imponderables aren’t being addressed. That’s not too encouraging, because the greatest art is basically about imponderables.
In wary fairness
Before I rip a few rather large holes in this rather complacent view of future art, I want to be fair to the basic idea, particularly the software side:
1. Art software, notably Adobe Illustrator, Corel and other high-end software, is incredibly useful. It reduces the “nitpicking factor” in a vast range of practical ways. It’s surprisingly efficient, reducing time wastage, and allows a lot of possible options for using even individual elements of works of art.
2. This software has created a “language” which is far more comprehensible to outsiders than the very esoteric and often introverted language of fine art. Anyone with basic media knowledge can follow the production logic of a logo, for example. Unintentional as it is, this is a very useful way of communicating and explaining issues to non-artists.
3. Art software can be almost incredibly productive and deliver practical help with improving efficiencies when doing major volumes of artwork. You can get a lot of the grunt work done, and stay on track with working at your own intellectual speed, which for artists can be faster than light, quite literally. The manual work goes much more smoothly.
4. Cognitive computing art can potentially help with getting the software to do things it wasn’t designed to do. Basic software has limitations, and those limitations can be trip hazards for artists.
Cognitive computing art – Proof required
OK, that’s the fairness taken care of; now some less enthusiastic views. Cognitive computing is likely to be the big thing for many decades in software. It can solve problems; show how it rewrote code to do things, and much more.
That doesn’t instantly mean cognitive software can produce great art. Most current computer generated art is a combination of fractals, Mandelbrots, and other well-known cookie cutter stuff. As wallpaper, it’s great. As art, it’s nothing much. It expresses nothing, very blandly.
Another issue, and this is almost computer heresy, is that anyone who’s ever used art software can tell you that it’s a virtual recipe for heavy use of the Undo button. Some things computer software can do with images just don’t work; they’re essentially useless.
The tech side of this is that computer software works in degrees. X software function + still life usually just means a messy image. Would have been great for Expressionism and other “spare paint” work, but pretty pointless for line work.
The “dazzle factor” in computer generated art imagery doesn’t often survive actual analysis. Sure, it looks good, but art? Barely, if at all. For ignoramuses, it’s that splash of color to go with a lost peasant vocabulary and a very undemanding audience. For artists, it’s a recipe for a dripping tap of compulsive imagery.
If you check out Dharma Wheel images of computer generated art, you’ll see a very mixed bag. Some of it, notably Dharma Wheel’s formulaic and very well colored example above, really is quite fascinating, but you can practically see the formula. As a reason to take up mathematics, it’s great; as a reason to become an artist, it’s debatable. “Generated art” is described by Dharma Wheel as:
…a work of art that is generated in a random and automated way by a computer program with the use of mathematical algorithm. For an artwork to be considered generative art it must have the influence of the artist, no matter how limited. The artists sets the ground rules for the formula before the random process can take over.
Nice to know artists are the drivers, not some five-minute wonder of algorithms. Admittedly, this stuff isn’t “cognitive” computing. It’s formula-generated. The software hasn’t learned this, it’s just reproducing its code. Some people have called this “neo-psychedelia,” which is almost right, but the real psychedelia habitually went off the map artistically.
The exact line between lots of drugs and lots of inspiration is open to debate, but inspiration is very much about the human factor in art. Could cognitive software produce a Van Gogh? Technically, yes. Artistically, no.
Software can do intricate Baroque or Rococo with ease. Character, no. It’s the difference between décor and desire.
Oliver Brown, a post-doctoral fellow at the University of New South Wales, wrote an interesting article in the Sydney Morning Herald regarding cognitive computer art, covering the “usefulness of art” in some depth. The practical economic values of art are often ignored.
Usually, I’d ignore the extremely annoying, trite, “usefulness of art” as a typical sink hole devaluation of human experience by people who hide behind external rationales for everything. In the same mindless perspective — what use is happiness? What use is a baby? What use is beauty? What use is a smile from a friend? What use are you?
Brown redeems himself almost instantly with this quote:
“ …possibly the best take-home message for thinking about the future of computational creativity is to do as the artist does, and be prepared to be challenged: to rethink your assumptions and your understanding of what is fundamental to human nature.”
Quite right — screw the rationales; art is about people’s realities. This drab, utilitarian view of art as some damn slave to “usefulness” is for machines and herd-thinkers, not people. That said, when it comes to equating cognitive software with usefulness, it’s an interesting addition to the argument.
Let’s talk uses — in fact, every damn manufactured thing you’ve ever owned or seen is based on some form of art, from the engineering design, which is a type of art, to the presentation. It’s that practical, and that useful.
Now let’s consider a real designer lifestyle using cognitive computing — imagine a 3D-printed Gothic toaster. What would it look like? Would it come with a built-in robot maniac playing a steam organ or not? If so, how much would it cost? See any clichés and snivelling utilitarianism in that mix? If cognitive art can free up design in context with the new tech which will rewrite human lifestyles, it can be a positive — provided you don’t muck it up with “usefulness” and other ancient-hack ideas.
That has nothing to do with art in real life. Art is always a major part of your preferred environment. It’s a type of indirect expression. Forget the bland, utilitarian lifestyle approach and focus on preferences for a second. They’re totally different things.
It think cognitive art will turn in to a way of creating personal environments, harnessing the practical side of a computer program which learns to actual needs. You’re more likely to wind up with a design-based personal space, using cognitive software as the machinery.
Imagine a life where you can use cognitive art and design software to create your personal heaven. Not some half-ass media exercise. Art can ignore limits. It can go where the mind can go; if cognitive software can help, great.
A bit of mongrel code for cognitive art:
*List X
*Let X equal unknown imagery.
*Read anything that’s not banal, or reinforces mental drudgery.
Yes, you can escape the predictable, whenever you want. So do it.
