The Future vs. Art

The ArtMoi Podcast

or Listen via Stitcher

Episode 2

Machine Learning

Expert: Artist and Computer Programmer, Jeremy Rotsztain

M: Greetings art world futurists!

Welcome to the second episode of The Future vs. Art, an experimental podcast series where I attempt to future-proof the art world by testing out a few scenarios.

To do this I investigate tech trends, and imagine how these trends could impact the art world in the future. This may prepare us all for winning or losing against technology, hence the future vs. art!

My name is Michelle Bouchard, and I will be your host/time-traveling enthusiast on this journey into the future.

Content for this Podcast is produced by ArtMoi, the catalogue and art documentation platform for artists, galleries and art collectors.

ArtMoi is working to create a new standard in the industry that is controlled by artists for the long term provenance tracking of art.

To learn more, don’t forget to follow ArtMoi on Twitter, Instagram and Subscribe to the ArtMoi Blog.

In last week’s episode, we talked about how cryptocurrency and blockchain might actually make the art world safer and more efficient.

This week we move towards humanless interaction and machine-learning.

I personally feel self-sufficient without one, but acknowledge that Personal Assistants like Alexa by Amazon, and Google Home, are a huge craze right now. They have machine learning capabilities that can anticipate the decisions we make..

They get hooked up to the cloud, our wifi, our bank accounts, and get to know us and our preferences faster than the average human. From personal experience, thinking about a future in which machines think and act for us can be shocking, and a little stressful.

What if Personal Assistants become so advanced that they anticipate much more than buying our groceries each week, they create our art?

We talk an artist and computer programmer Jeremy Rotsztain who weighs in on this.

First things first, I’m going to put it out there that machine learning is a field of computer science that gives computers the ability to learn without being explicitly programmed.

It is an application of Artificial Intelligence which gives machines access to data that lets them learn tasks and answer questions.

For example ‘Google...what’s the weather today...Alexa...I need organic tomatoes from Whole Foods…’.

You can be conversational with these machines and as a result they learn more and more about you.

In terms of creating art and machine-learning, this has been going on since the 1970’s with the first printers transforming image information into a variety of new media art including languages, alphabets and even music.

So in the future, what is at stake for artistic integrity, if personal assistants can be trained not just to perform but to create?

Without further ado, let’s jump 60 years in the future, and imagine a machine learning art take over.

The year is 2078.

There is a large debate on promoting the artistic integrity of machines, and personal assistants because they not only perform basic tasks to complete your grocery list, but also create a custom style of art.

Machines have developed an understanding of art that is superior with more perfect results for the individual.

There are also machines that can learn to produce collections of 1000s styles, subjects, and mediums, more art than anyone artist could create in a lifetime.

Need art for your new apartment without having to leave your apartment? Well, ask you personal assistant.

A significant part of the art world is now suffering as machines produce original artwork on a massive scale beyond our wildest imagination.

I came back from this machine learning art take over, and joined Jeremy on Skype from Portland Oregon..

Hi Jeremy!

J: Hey Michelle, how’s it going?

M: Good! Nice to see you, okay you can’t see me here we go...

In hindsight, I probably had too many questions with the future of art being at stake and all, but I had to find out how Jeremy does it all.

episode coverart

VR projects – House of Shadow Silence & Ascension (@Upfor)

M: I saw your website, and I jumped at it because you do such incredible work with machine learning and digital art and you can also write algorithms if I’m not mistaken.

J: Yeah I’m a coder by trade and it makes it way in all the artwork that I do.

M: Wow. So could you describe your body of work?

J: That's a good question. I'm a digital artist. I also sometimes get described as an algorithmic artist, or a generative artist. And what I do is, I write software to create imagery often using the visual language of abstraction. Historically, like geometric abstraction, and kind of, Colour Field painting and so on. I use it as a mirror of sorts to reflect on contemporary technology. So how cell-phones, or the web, VR and so on affect our lives, and our ways of thinking, or ways of behaving. Our day to day anxieties and things along those lines.

So my work kind of manifests itself as animations, moving image art so that includes video, which I sometimes present in kind of a gallery context. I've been doing a lot of work in the last year in virtual reality, which is a great space for algorithmic art. I've also made iPhone apps, lots of digital prints as well. Code manifests itself as we know in this day in age in lots of different mediums. Not just the web, not just your phone but makes it way into real life, and so I found applications for lots of different media along the way.

M: That's incredible! You have such a huge range of art, and you cover anything imaginable in terms of technology and resources. So do you find it hard to translate that into typical artwork which is usually some sort of narrative?

J: Sure, sure. I'm definitely inspired by artists who have worked before me. I definitely take inspiration from previous techniques or making work, not necessarily the same stories that they told but I guess, you know, maybe using the visual language that an artist would have had before, let's say Rothko, and to kind of tweak his ideas about colour and space into something that's more contemporary.

Some media artists are very, you know, just purely forward thinking, and are historical in their approach, and I probably sit in the middle. Some of my work is inspired by ideas that come across in reading art history books and so on.

M: You mention Art History, and when I think of Art History I think of something super traditional and Eurocentric - really old ways of thinking and so when we use technology to create art I expect that we run into some problems in opinion, and we run into some skepticism.

episode coverart

VR projects – House of Shadow Silence & Ascension (@Upfor)

If we flash forward into the future when technology totally takes over our lives and is completely integrated, if personal assistants could create artworks, that could create some big conflicts. Do you foresee that happening in the future?

J: I guess it really depends on your attitude towards what makes art, I guess. Meaning, like, it could be some artificial intelligence algorithm where you can give it some set of instructions, or idea that you're look for, you know - Mix Monet with McDonalds and it comes up with whatever that resulting work is. I think that's still very much part of this idea that's guiding the work.

I personally, as an artist and programmer, I actually really find it so meaningful to have a deep hand in making the tools that I use so to write code to generate images that I don't know if I'll let go in that same way. One could say that the computer is doing the work but I still feel that humans have a strong hand in machine learning. Like a studio model they're guiding what that work is. So if you have an artist, who - is a prodigy, maybe you can think about the machine learning as the artist prodigy.

M: So. Right now it sounds like the artist is in control of the machine.

Let's just think of a scenario in the future where we've programmed machines to create on their own. Is that possible? What would artistic integrity become of that?

J: That's a good question. So, it makes me wonder who's looking at the work and who are they making it for?

I still feel like so long as there is a coder or designer behind the system then it's influenced behind the ideas of the designer. So, I find it hard to buy into this idea that the work is a machine work on it's own merit.

Does that make sense?

M: Yeah.

J: That there's always someone programming the system with a set of values. Like, machine learning is based on previous information. Right? It takes a body of work, a body of images to decide what to do next. And so by deciding what you feed into the machine system, from the beginning, you're influencing the outcome of the machine's way of thinking.

So I still feel like it's hard to separate the machine from the human creator, and the designer. So, I don't know if it can be on it's own.

There's all sorts of interesting tests that have been done along those lines. Can a machine make work? And I guess, I wonder, you know, if it's a visual that’s just enough to make it art, and if there is something else, beyond the novelty of the machine making the work, does it make it art?

I also wonder if, maybe, machines will make art for themselves, and therefore it will be it will be art on it's own. Maybe it's not even for humans to appreciate but for the machines to share amongst themselves.

M: A machine learning community of machines creating art. That could be an interesting future!

J: Precisely! I mean it's out of our own understanding of what art is, like a whole other language.

M: Totally! In your experience what do machines like and what do machines not like, is that a thing?

J: I guess it's all personal interpretation in terms of better or worse. I can't remember the name of the study off hand but there's been a bunch of interesting studies that have used AI to create work and put it infant of an audience, and that audience not only found good looking work, but they also liked some of the work more than other pieces.

I think it was all purely abstract that looked like Colour Field painting, but they deemed that it looked like it was real but also more interesting.

I should point out that I'm actually not an expert on AI but it's definitely of interest to me. It's a more advanced way of making work that I'm involved in. And for me, personally, as an artist I'm still looking a lot at making abstract form and I have a lot of hand in creating and designing those geometries - coming up with the mathematical formula for generating 2D or 3D shapes I guess in my own work I'm looking for a kind of visual richness that feels a little bit more, kind of like the natural world, it feels a little bit more analog.

So not exactly hard edge abstraction, or hard edge forms but things that are a bit more visually expressive. I guess that's my own kind of approach to technology, is to give it a kind of visual richness that earlier, primitive geometry, or of earlier 3D work may not have had.

M: Would you see any benefit to machines having a hand in picking -

J: In picking the best options?

M: Yeah exactly.

episode coverart

Electric Fields II - Electric fields and Action Painting

J: I guess. I wouldn't do it because I wouldn't feel as much of a hand in the final work, you know? I guess as a programmer I'm a control freak, maybe that doesn't surprise you, but yeah, I still like to have it. I still like to be involved in the decision making along the way. I think that people still value that there is still some artist judgement in the process.

M: There is a real value to have a relationship with a human being who created something rather than a machine that's programmed to create, I guess, an ideal, perfect, visualization of a form of art.

J: Exactly. I think it's missing the kind of human element in it all.

M: So, art, hopefully, will continue to be human in future.

J: I guess what I'm thinking about is this question of whether will you find the work to be more meaningful if it's described to you by a database or a data-driven system that maybe doesn't know the history of the artist or has not purpose or connection, versus coming to a gallery.

I don't know! I guess it's really a person by person basis, or a buyer by buyer basis if that would be a kind of meaningful experience to them but -

M: Yeah, I'm not quite sure either and I think building rapport, building relationships, with artists or gallerists, or dealers will always be a thing. I don't think it will matter 40 years into the future because that's what protects the art world and that's what it thrives off of: relationships.

How have people responded to your work?

J: I show my work in a lot of places. Online and in physical space. I guess I'll make larger project that I'll show at festivals etc. In between I'll kind of work on little sketches where I'll show more on social media and the like. So I guess I have a few different audiences for my work.

I've done some site specific installations using virtual reality and that's what I've been focused on in the last year or two. So it's a little more bringing the digital into the physical space and having a strong interplay between animated content, animated forms I should say that exist within a sort of simulation of the actual physical space the viewers are inhabiting.

And those are a lot of fun to kind of turn peoples experience of physical space on its head by showing them a virtual copy of the world that they're existing in that has all these kind of impossible forms. Where time is slowed and turned upside down that's what I've been enjoying about a kind of viewer interaction so there's a kind of play where reality becomes - After the experience the reality would appear to be a little more plastic and valuable then experienced before walking into the show...

M: Do you know any artists that are really interested in becoming exactly what you are: an artist who creates VR and programmes computers to create visual art?

episode coverart

Gunfire - Electric fields and Action Painting

J: Certainly! I think as people group online and are embedded in their phones, art is inherently digital for a lot of these people and so they're looking to work more and more in this space. I think a lot of schools are actually pretty behind in the process and so you see new models for art making pop up there's a school for poetic computing run by one of my mentors Zach Lieberman in Brooklyn teaching digital art practices using programming algorithms for making images and doing artificial intelligence and art with interaction.

There are a lot of new models. The grad programme I went to actually is old school and ground breaking at the same time. It's been around since 1981 it's called the Interactive Telecommunications Programme, or ITP as it's well known and it's also in New York, in Brooklyn.

I guess it's new logo or tagline is 'The School for the Recently possible'. And they're just trying to innovate at steps along the way, kind of playing with new technologies and seeing what the human potential for all of these - human/artistic potentials for all of these are part of all these new technologies as they come up so as the iPhone came up people were on it, and as VR kind of became affordable students and teachers have jumped on it, there's a hand full of specific schools with new media art programmes that are innovative along those lines.

M: So we're seeing it already which is super exciting!

J: Totally, and it's nice to see what fresh eyes bring to these new platforms.

M: Yeah, so what's next for you?

J: I guess I'm still pretty keen on VR. I feel like there's - VR is in its second, maybe third iteration and the technology has pushed forward so much...

For the rest of the interview, Jeremy’s projects and programming abilities continued to blow my mind.

His latest work deals with Virtual Reality, and a type of Mixed Reality that projects images onto surfaces as well as HoloLens’. All things I wish I had time to get into right now!

The good news is, especially for those skeptics out there, is that art wins this week. We know that this type of technology might be possible, and whether it has merit is still not clear.

Also, I’ve taken away something that warms my art loving heart in spite of how the cold this technology may seem.

Although Jeremy is working on the cutting edge of machine-learning technology to create a huge range of art. The bottom line is that humans, for now at least, have the upper hand.

Although machines are well aware of what they are doing, they are merely guides where artists are the idea makers.

They can learn, but they need a master. And for now artists have control.

That’s The Future vs. Art this week, thank you for listening!