What actually is an "AI Generated" Scultpure - Process

I saw a video and some photos of this sculpture come across the NY Times instagram feed yesterday.

Unfortunately, I didn’t see the image(s) of the work on it’s own. I only saw the image with the text below:

A curious title! What does it mean?

I think we are already starting from muddy water here.

Let’s fill in with a few more images and text to get a better idea of WHAT ACTUALLY IS going on here!?

Where are we starting from?

What is the process?

What has been made?

Are we talking about what is made or of the process?

OK! Cards on the table on process.

  1. An MIT based artist did a residency with corporate sponsorship of tools and resources - Nothing new there, happens all the time. Getting a visual artist to use a new tool, tech, and then making content about/around it, is basically the creative foundation of marketing and advertising work. How do we take a thing we’ve made and communicate it to other people in a way that makes it seem relevant/desirable, etc.

  2. The artist primarily used the new text-to-video tool - that’s the AI tech used.

  3. The AI tech generated code - that through different interfaces, that code could be used to “make” or print things - using other existing tech. So the human artist starts with a prompt (human energy) and then also connects different medium making tools that already exist to interface with this code.

  4. The labor that AI performed was to produce some code based on prompts given by the artist, and then that code could be used to interface with existing technologies.

Therefore, the AI, did very little in the sense of what we would call creative generation, and much more in the labor or code generation.

We don’t know if the software produced hundreds of images, and the artist just picked their favorite - we just have what we have, and the info provided.

Let’s take this marble form to start.
Let’s talk about the actual process.

If a human artist wanted to make this form from start to finish it would take hundreds and hundreds of hours of physically demanding and highly trained labor to simply fabricate this piece. Not even talking about the idea of the form - just the labor to make that form would be almost impossible. So before we even talk about what AI did, perhaps the focus should be on the robotic arm that actually made the thing.

So the actually labor of making a marble sculpture is something that was done, and is done by human labor all over the world.

Human labor that is connected to an economic system that turns mountains into useful products.

AI is not replacing human labor - the 6-axis robotic ar,, is.

AI just “had” an idea, and generated some code.

The AI - text to video software - did the least amount of work here.

Not unlike how ideas work in our brain.

Ideas are free. We have them all the time.

For those of us that can visualize things in our head - that IS text-to-image/video software that you already have in your body.

Your ability to imagine an abstracted human-life form is already within your operating system. It comes pre-loaded to your OS.

Many of you will think of a form right now, just by reading these words.

That is the opener of the article (above).

So many assumptions already out of the gate.

  1. ‘Alien-like torso” - why is it an alien like torso? Looks like a human form. If you saw this form in the distance, in silhouette, fully clothed, it would just look like a human form. The fact that we so quickly jump to “alien-like torso” is because artists before have made drawings, sculptures, and movies of aliens that have a torsos like this. That’s art. That’s not nature. That’s text to image technology executed through skilled labor. We have those artifacts - and those artifacts can tell us something about ourselves. Why do alien like torsos just look like human torsos with exaggerated proportions is a better question, I think.

    Further, do those exaggerations say something?

  2. Yes - an image has become an object. But that image (or lines of code) did not become an object without a labor partner - someone to channel the image/idea through a medium. In this case, a very powerful machine was used for labor.

Ideas are free.

It seems that how the AI is functioning here, is to actually make that statement fully true.

If ideas are free - generated in our minds all the time - then the work that AI is doing is preparing those ideas to go out into the world and plug-in to other interfaces or mediums to produce results in the physical world.

That was clunky.

The tool of this AI, it seems, is to speed up the pace for communication between interfaces.

So if an idea can become code, and that code can be generated faster, then we can make a new thing, better thing, or just more things faster.

Faster is good for things that need to be faster.

Not all things need to be faster.

Now what?

We have the thing. We made the thing faster than we could have before.

Now what? We have the thing.

It’s here now.

If you or I came across this sculpture in real life, is there any possible way we could know that the idea came from a -text-to-video prompt?

Is there any possible way we could know this process just by looking at the thing?

If we didn’t have a text prompt how would we even start to interpret the work?

Is there any kind of art that didn’t start this way?

It’s called imagination.

AI, in this application is an “Imaginator.”

But someone has to hook up the other tools to the imaginator.

Just like we have to use tools to partner with our imaginations too.

This process isn’t really anything special - we all have the ability to imagine so much - it’s infinite.

That knowledge. It’s functionally infinite. It’s always expanding.

Wisdom is different though.

Wisdom says, ok, so we’ve got the knowledge, now what do we do with it?

We can do anything, so what do we do?

And when we evaluate what we’ve done, let’s look at the physical result of what we’ve done, rather than just the process.

Next post will be about the sculpture itself - and how to read the work itself.

I pulled a Modern Art HIstory book off the shelf to look up some other human-life sculptures.

Who knows, these could’ve been AI generated images too!


Brancusi, 1913


Published on by Eric Trine.