9 Comments

I'm luddite enough to be frightened by it all. I keep looking for the pony in the room full of shite, but inside, I'm just....scooping. We are in uncharted territory. It's the antithesis to the computer in Star Trek: TNG that created food and beverage, saving mankind, re-ordering the economic values in such a positive way that all benefited from it. Instead, we get the destruction of capital, the pretense to art and an investment in things that create images of our past--while our future is a desert landscape of creation and innovation.

Expand full comment
author

The point here, Mark, is that while you may be a luddite, the process of creating sophisticated, print-worthy, compelling art is so simple that even a Luddite could excel at it. The reasons I would ever spend money on a commissioned piece of art now are reduced to very few. And I emphasize again that the technology, despite being very good, is in its infancy. It's going to become exponentially better.

Expand full comment
May 16, 2023Liked by Tom Wark

"Pictures worth a thousand words..." Images illustrate. This piece offers one of the clearest and most understandable illustrations I've seen yet about one tiny aspect of "AI" -- its promise and its threats. Nice job ! Thanks ...

Expand full comment

Remarkable! Loved the drawings the computer did. The trend of rewarding the super talented (developers of AI, for example) and penalizing the mediocre (most of the rest of us)continues. Great job, Tom, once again.

Dave Jefferson

Expand full comment

The sticking point with AI, though, is that it can't innovate. There will never be a new style of art, a new technique, without humans first breaking that ground for AI to "learn" from. If we hadn't already established "art deco" and "watercolor" styles, AI couldn't give you anything like it. Maybe the technology will get to that point, but nothing in the current AI models allow for out-of-the-box "thinking". If what you want is something you've seen before, AI is your go-to. If you want something that stands out or is different, I think AI is still very far off from that, if what we currently call AI even can go there.

Expand full comment
author

I don’t think we are far away at all. Insight is the ability to make connections among disparate ideas and facts. It can do that. And it will only get better. It will always be a tool because it has no soul and is not mortal. But the consequence of the tool is to diminish, devalue and suppress a kind of human creativity that has sustained commerce and livelihoods for millennia.

Expand full comment

"Insight" isn't creativity, though. Insight is understanding, it's educational at best. It's ingenuity, but not creativity. Creativity (especially artistic) is finding something new, that affects a large number of people, for reasons difficult to put into words. What there is still zero evidence of, so far, is that AI can create something new without already-established context for every single element. Not that humans create in a contextual vacuum, but when we synthesize, the breakthroughs are when the sum is (somehow) greater than the parts. AI is simply very good and piecing a puzzle together of exactly the pieces asked for. I can see why corporations and marketers see a lot of potential in this. But it isn't actual creativity, it's just very, very good replication. I've yet to see any examples of AI product that isn't pure replication of precisely the elements asked for.

Expand full comment
author

That's a good breakdown! I'm slightly annoyed that the article credits GAN AI as being able to do "transformational" creativity, aka entirely new styles, but then doesn't have or link to any examples, and only posts the usual prompt-based imagery of more limited AI. It does make a great point though that AI can't (yet) really understand the relationship between things in an ironic or poetic way, it can only produce very matter-of-fact representations (for lack of a better way of typing that).

Expand full comment