In the creative world of Joanne Hastie

Explorations of AI Art— Episode 05

Beth Jochim, Director and Co-Founder at Cueva Gallery. Twitter: @_bblurred

“Combining the two worlds — painting and programming has been energizing to my art practice.” — Joanne Hastie

Robot Painting no 130, Painted by hand + a robotic arm. Courtesy of Joanne Hastie.

After a short summer break, Libre AI is resuming the interviews with artists that use Artificial Intelligence and Machine Learning techniques in their practice. This time, I had the pleasure to interview Joanne Hastie, a multi-disciplinary artist based in Vancouver, Canada. She combines programming, robotics and paint to create art and be more creative. Her latest work of Machine Learning Curation was shown at NeurIPS 2018 conference in Montreal, Canada and at the 2019 Women in Data Science Conference in Zurich Switzerland.

Let’s discover together a little more about Joanne’s work and approach to art.

Beth Jochim: Joanne, you are an artist interested in experimenting. How does your emotional and conceptual participation change in the process of making art, using different medium and techniques (from hand painting to coding to robotic arms)?

Joanne Hastie: I’ve been painting for over 10 years before I started coding my art. I painted brightly colored city scenes with discreet brush strokes. I took on the challenge of painting with code and a robotic arm by examining my own painting process. For example the robot arm cleans the brush between colors, just as I do when I am painting to get discreet colors.

There are so many uneventful tasks within the painting process that an artists does not realize they are doing. To automate the painting process, I had to become aware of them and find consistent patterns that could be programmed. For instance:

  • Why I select colors and in what order
  • How to I apply paint
  • In an area and with each brush stroke
  • When and how often I clean my brush
  • When do dip my brush to get more paint
  • How does each brush stroke look between each step

So now rather than being focused on creating a final image, I am focused on the process itself. So my participation has actually changed but my technique is the same. Maybe the viewer expects me to be farther away from the process using robotics and automation, but I am significantly closer to the details now.

Joanne Hastie and her robotic arm whilst painting abstracts. Courtesy of Joanne Hastie.

Beth J.: On your website we can read that you are interested in using ML/AI to be more creative. Indeed, you say this technology allows you “to create more ideas faster, to have a better understanding of your artistic preferences and to create better art”. It seems AI is acting as a sort of extended mind, could you elaborate on this concept for us?

Joanne: Only recently, I have spent more time reading blogs and watching videos about generative art to improve my programming skills. I immediately noted that opportunities described for generative art are the iterations that can be created. When I started using code with my painting process, I immediately thought this was novel because as a human painter I had limited ideas when I approach the canvas or page. So perhaps my imagination coming from a fine-art painting perspective, then investigating generative art I am inspired to create multiple paintings from one single idea or block of code. So combining the two worlds — painting and programming has been energizing to my art practice.

Often a painter will stick to paints and a color palette they are familiar with. It takes years to get comfortable using mediums and paints that produce expected results, for example paints dry different colors than when they are wet and each color has a different translucency. Often a painter pick either oil, acrylic or watercolor, and stays with that medium for years — different tools and different workability.

Prototype IX. The curves made by the robot simulate the fluid motions of human hand. Courtesy of Joanne Hastie.

When programming a robot, I have not limited the color selection in the code based on my favorite color combinations to paint with — instead I mix the colors based on the digitally generated image. Right now I am working on an algorithm that compares the digital to analog versions, I hope to improve color selection with future compositions. But overall I am experimenting a more color variety than I would if I was an artist first experimenting with abstract paintings as I am. However, in contrast, there are limitations; my robots have limited number of paint containers, that, in turn, limit the “number of colors”. Most artists have a palette with a range of colors mixed rather than discrete colors. Mixing a range of colors with a robot is a challenge I’d love to tackle in the future as I continue to program my painting process.

Beth J.: What do you think is the added value of AI Art for the public? Do you think that AI art can give the spectators a different experience compared to more traditional art?

Joanne: I see the opportunity for AI Art for the public is to connect with the art and artists in new ways.

Early on in my engineering career I was inspired by automation and reducing tasks down to their basic steps. As an intern, I was able to design semi-automated equipment to assemble medical devices. I found it captivating to work on the assembly line and examine motions with our hands and then try to mimic it with contraptions.

I love breaking down the process, so for me applying it to my painting practice has been amazing as I can share this passion with the public. Whether I am explaining how a GAN placed my signature hints of red paint into the image or how the robot placed a brush stroke, I am able to articulate my artistic process in a different way and connect with people on new ideas. When I was working in automation, my work was not available or visible to the public, so it is inspiring to show my work. Same goes for painting, it is not easy to hand-paint with an audience as it is a very personal process — so painting with a robot I can create experiences.

I also believe art can push the technologies further in more variety of ways as there are so many ways artists can apply these technologies to their work. With artists sharing what they are up to and code being accessible — it is an exciting time for technology and art.

A table, a robotic arm and a human hand. Courtesy of Joanne Hastie.

Beth J.: Is AI Art emotional and able to amplify the response of the public, or is the experience only for the artist invested in the process of making a new artwork?

Joanne: In any art genre, I think, it depends on the art and the viewer as to whether there will be an emotional connection.

However, because of our preconceived notions of technology, I think the viewer will bring their bias into the experience of looking at AI art. So there is an opportunity, as well as a challenge for an artist to consider the viewer’s expectations. Because of this, my initial series with the robot is loose, varied brush strokes. I wanted to work on paintings that the viewer might normally expect to see done by a human hand rather than an accurate, repeatable robot. For example, I spent time programming the robot to paint curves rather than straight lines, as a human artist using fluid motions often paints with a wrist or elbow motion inducing a curve.

Beth J.: Is there a question nobody ask and you would like to answer?

Joanne: In a past engineering work, almost a decade ago, I was designing mountain bike frames. In two different projects, I created hundreds of versions of the designs and then manually in Excel used data analysis to find correlations in the individual features from my 3D models to optimize the structure of the frame and suspension design respectively. These projects took months iterating over hundreds of designs and then finding the correlation in the features to come up with the best designs within the constraints. I strongly felt the designs benefited from the rigorous review.

Having those experiences and now learning the latest tools in machine learning that automate the analysis process has been absolutely energizing! I very much enjoy the discovery process of iterating through a project, now I am automating a thought process!

So although it is less visible than a robot making marks on a page with a paint brush, I think this past experience has made machine learning more approachable for me.


“The simple truth is that companies can achieve the largest boosts in performance when humans and machines work together as allies, not adversaries, in order to take advantage of each other’s complementary strengths.”

― Paul R. Daugherty, “Human + Machine: Reimagining Work in the Age of AI”



This quote from Paul Daugherty is thought for the business, but I like to bring something home from what he says and apply it to the art scene.

Art and technology do not have to be in contrast or mutual exclusive, but they can work together towards creativity augmentation. Collaboration and mutual inspiration is what comes out also from the chat with Joanne.

The use of technology has allowed her not only to focus more on the process of producing art, but also to think differently and create multiple paintings from one single idea (or block of code).

She has got new insights about her personal subjective preferences, she has overcome the limits of her favorite palette incorporating in her work a bigger variety of colors, and reached loose abstract painting through robot randomness.

Biases from the viewers side are expected, but approached in a way that allows her to embrace a challenge and an opportunity at the same time.

She has learnt to curate and classify her compositions, getting new insights for future works.

Using a robot arm to paint gives her also the great possibility to connect and interact with the public at a level that was unimaginable before, creating a shared experience for the audience and a flow of new ideas to bring into new projects.

A big thanks to Joanne Hastie for opening her experience and thoughts to us. Her evolution from a realism painter to a technology artist is truly inspiring and confirms that this is an exciting moment for the art scene and for artists who want to achieve new results in collaboration with technology. ∎

About the author: Beth Jochim is the Creative AI Lead at Libre AI, and Director and Co-Founder at Cueva Gallery. She works at the intersection of technology and arts. She is actively involved in different activities that aim to democratize the field of Artificial Intelligence and Machine Learning, bringing the benefits of AI/ML to a larger audience. Connect whit Beth in LinkedIn or Twitter.

Creative Commons License
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.