#IamthefutureofAI Series: Lia Coleman
Artist, educator, and AI researcher, Lia Coleman takes us through the highlights of their eye-opening career journey including the experiences which inspired them to get involved in the space, the challenges they’ve encountered as they navigate the world of technology as a creative along with the importance of diversity in the space and having allies advocating for you, helping you to push forward.
This interview is part of Women in AI Ethics (WAIE)’s “I am the future of AI” campaign launched with support from Ford Foundation and Omidyar Network to showcase multidisciplinary talent in this space by featuring career journeys and work of women as well as non-binary folks from diverse backgrounds building the future of AI. By raising awareness about the different pathways into AI and making it more accessible, this campaign aspires to inspire participation from historically underrepresented groups for a more equitable and ethical tech future.
You can listen to the podcast or read through their conversation below.
What really inspired you to join this space?
Yeah, for my professional life, one thing that happened that inspired me to join the AI ethics space and really value diversity and having people from a non-dominant demographic at the table, creating algorithmic decisions. Yeah, my first job out of college was working at Facebook and I was training these image and video classifiers to take down bad content.
We had a variety of different spaces that we had worked on as a team like suicide, self-harm prevention, child pornography, graphic violence, terrorist propaganda, white supremacist content. A wide variety of really, really heinous reprehensible things in society that get posted to the platform.
And one thing that was really eye-opening was that I was the only female software engineer on my team. And when the topic of deciding who is going to work on what vertical, what focus area came up, I was pretty much the only person that wanted to work on child pornography because I was like, “yeah, that’s really, really important.”
And it did feel like a lot of other people on the team were not excited about it or not wanting to work on it probably because it is a very heavy topic. But I do feel that part of my motivation to work on it, it was probably influenced by me being a woman and thinking that it’s really important to make sure that kids don’t get exploited in this way.
Whereas, possibly — this is just speculation — but possibly my male counterparts on my team didn’t realize or think that it was important or as important. Maybe it’s some mothering instincts within me to protect these kids. This is a really important issue that we need manpower behind, we need engineers to be working on this space.
Whereas if I had not raised my hand and volunteered to work on that nobody would have worked on it. Working on that issue of child pornography, it was really eye-opening because the team that I was collaborating with — I was part of the technology engineering team trying to train or create these machine learning classifiers to recognize child pornography and flag it.
And then the team that we are collaborating with, they were the actual content reviewers for people that would be contacting the government of local municipalities and flagging the police for specific things. And that team, it was interesting because that entire team was comprised of women. And it was the first time I was at Facebook and I was working with so many women as coworkers. And it was just really, really — it was a breath of fresh air. It was so nice.
And I think it is possibly something that has to do with gender then that makes us care for children and work on these. While they are really heavy, serious, and can be traumatic at times to work on these areas, it is really, really important for society and the protection of these children. That might not be as top of mind for our male counterparts. So that’s why I think it’s really important, circling back to why I’m in the AI Ethics space. That’s why it’s really important when you’re creating AI technologies to have a really diverse set of opinions at the table. So that important issues like child pornography are worked on and not swept to the side because no engineer on the team wants to work on it.
And secondly, because I wasn’t super clear on what this space meant, I just talked about the AI ethics space. But then also within the specific space that I focus on is AI art creation as an artist who creates my artwork with AI and coding — programming. And this is a much, much lighter anecdote.
When I was a kid, I was in sixth grade. I was a part of this Women in Engineering Summer program. I guess this ties in with the first anecdote related to gender and being a woman in these tech spaces.
When I was in sixth grade, I — always as a kid, really loved to draw, I really loved art, creating art. And also in school was told because I was good at math, good at science, I kind of took the programming technical person route. That’s also due to the encouragement from my Chinese American immigrant parents, “be an engineer”.
In sixth grade for the summer, as part of these women in engineering-type program, we took a field trip to the MIT media lab where it was a group of 10 sixth-grade girls. We show up at the media lab. There was this woman there. She heads this research group at the media lab on computational textiles. And we did this workshop with her, where she has this table full of different scraps of felt, LEDs, conductive thread, and needles. And she shows us how to make these bracelets, where we took the conductive thread, sewed together a circuit embedded LEDs into the felt strip, and then made the circuit where once you snap the bracelet shut, the circuit completes, and then your LEDs light up. And it was mind-blowing to me as a sixth-grader. Because up until then, I had thought that art media and technical programming, they have to be separate people. Those have to be separate worlds. And I was super, super excited! I ran home and showed my dad this bracelet. And I was like, “Look at this! I made this!” I was really, really excited to see this first instance of art and creativity and design and crafting the mesh together with technology and electronics, conductor thread, and having those intimately meshed together. That was really, really exciting for me to feel that I don’t have to push my artistic side away when I’ve worked in a technical space. That was the first instance where I was like, “Wow, this is actually possible!”
It was actually possible to marry my art half with my technical half and to do work and have a career in this space. And that’s what I’ve been doing for the past couple of years.
How did you land your current role?
My career journey leading up to where I am now, I guess right now I’m about to start doing research on Creative AI at Carnegie Mellon this fall.
I currently am an artist using AI to create my work and my career path up to here was — in the previous question I gave you a little picture of sixth-grade Lia, who was like “Oh man, art and programming can be together. And then it wasn’t until college that I started kind of seriously trying to mesh the two together and after graduating college — well, in college, I majored in computer science with a math minor and went to MIT — and after graduating college, I went to work at Facebook where I focused on machine learning. And then I was training image and video classifiers because as a visual person, I really wanted to do computer vision and not natural language processing.
So at Facebook, I was training these classifiers — machine learning classifiers to take down harmful content such as videos of graphic violence, terrorist propaganda, white supremacists content, child pornography, really, really heinous things that were on the platform and trying to take those down before any human had to see them.
And then I was at Facebook for about a year and a half, I left because I was feeling unfulfilled creatively. I realized that Art Lia needs to have a creative outlet or else she feels really, really unfulfilled. So I left Facebook, I went to the School for Poetic Computation in New York City. I highly, highly recommend it. it’s this artist-run school that focuses on arts and technology and programming and computation and ways to express yourself through coding and technology and ways to be creative with the use of new technology. That was pretty much the beginning, as I think, I fit into this world of creativity and computing.
And then a couple of things I’ve done in that world since in the past two-three years, where I’ve been involved with the NeurIPS Workshop on machine learning for creativity and design. That also was a pivotal moment, kind of like the conductor-thread-bracelet-sixth-grade moment where I went to this conference at NeurIPS and it was the first time I’d seen in a technical -academic machine learning research space that artists were being asked to come on stage and give their experience of using TensorFlow algorithms to generate the lyrics for their album, for example. And that was really, really cool. And that to me was another pivotal moment.
That was eye-opening. I was like “wow, these machine learning researchers actually care what these artists have to say and this is awesome.”
And yes, I’ve been involved with that workshop, I’m organizing it. But first I was a participant and presented posters there and then I have also done a lot of teaching.
I was at the Rhode Island School of Design teaching Machine Learning Artwork to RSC designers. Also, have been teaching these online classes with Eric Schultz and he’s a great resource when it comes to making art with AI. For the past two to three years I’ve been teaching workshops and then doing various projects here and there related to creative computation. I’m currently working on a Music Visualizer to convert audio sounds that musicians have created and automatically turn that audio into a video being generated.
Coming from a non-traditional or non-technical background, what barriers did you encounter and how did you overcome them?
When I was working in the tech industry, one barrier that I encountered — I think part of it has to do with the fact that the tech industry is really male-dominated. And so as a result of having a lot of men together, they have a certain culture and a certain way of communicating with each other. And so people who don’t communicate in that way might not be able to get through to people or the ways that they do communicate, might not be received or might be mistaken or misread as something else that they’re not communicating.
So I did find when I was working in the tech industry that weren’t really a great fit. I didn't feel that I fit in with my team. I was the only female software engineer. And I think, in terms of ways in which I’ve gotten through that is even in spaces where the people you vibe with are few and far between. Maybe, the male software engineer, Allan, who is a strong advocate for the voices of women on the team and will push back. I think it’s really special when you find someone who acknowledges that there is inherent bias in the workplace and is really conscious and aware of that and tries to interrogate the behaviors of people on the team and just think about, “Oh, did the manager make this decision because of maybe some unconscious biases”, and bringing that up.
And I think more and more, I’ve built up a community of people and tried to surround myself with people that hold these views and are more aware and are not afraid to admit to themselves that they do home biases and consciously think about ways in which they can reduce these.
Why is more diversity — gender, race, orientation, socio-economic background, other — in the AI ethics space important? (Share specific benefits that more diversity brings to AI)
In terms of diversity and its value to AI and AI ethics, I think they’re inseparable. When it comes to ethics and the creative technologies that affect every member of society it’s really important to have representative groups or even voices of minorities at the table. So they can advocate and tell people their experiences of like, “The camera on my phone doesn’t recognize my face.”
And if there isn’t a person from that demographic at the table, that voice won’t be heard that experience won’t be — that story won’t be told. And then the creatives of the technology will never know like “Oh, for this specific segment of the population, our technology falls short.”
Which is why it’s super, super important to have people from all backgrounds and all demographics and all different genders, sexual orientation, race, ability or disability, age, nationality — all be able to weigh in and have their opinions and experiences heard.
Technology today and AI today is created by a specific segment of society, usually pretty privileged, white male, upper-middle-class to upper class. And these people are making decisions that affect literally the entire world in the papers that they’re publishing and the research problems that they’re choosing to take on. And it’s really, really important and I can’t stress it enough to have more diversity in this. The group of people that are making the decisions of what technology — what the next generation of technology is used for, what it’s created for, what it does because literally, the next wave of technology pervades the entire earth and all the people on it which is really important for how diverse the earth is, the human population is.
What is your advice to those from non-traditional backgrounds who want to do meaningful work in this space on how to overcome barriers like tech bro culture, lack of ethical funding/opportunities, etc.?
In summary, my first piece of advice was to not burn yourself out. To be kind to yourself, take breaks when you need to, tap in-tap out, allow others to step up when you’re feeling tired and allow yourself to rest because we’re here to make changes for the long-term. For long-term progress to be made, you do need to rest, take breaks, and be kind to yourself.
And my second piece of advice would be — if you’re from a non-traditional background to not — to try and create communities and find others who are from the non-traditional backgrounds. If there isn’t a space for you at that table, you can make that space. Especially, if the dominant culture is white, male software people.
When it comes to entering a space, allies are really, really key. As backwards and unfair as it is, sometimes people will only listen to people who look like them so that’s where allies come in.
When you find allies that are advocating for you to also not be afraid to ask them for specific things because that’s what they there for their allies. You can go like “hey, I’m looking to do this. Do you have anybody here? I’m trying to speak to somebody from X, do you know anybody or can you send me an intro email?”
Also, ask around. Even in spaces where seemingly it appears that everyone is homogenous. If you dig enough, often there’s at least one other person or two other people that are not of the dominant group and really, really see the value in what you’re trying to do and to sort of find other people like you, even if you are few and far between because I think it’s really important to have a team of people who will help you push this forward.
Because as I said, if it’s making the change in these spaces where the odds are stacked against you, it takes a team so it’s really important to find others who are trying to do the same things that you are. And from my personal experience, for example, I’ve been really, really pleasantly surprised again and again at the people I’ve been able to connect with and find and small little niche communities, I’ve been able to find because me, as a computer science major, trying to work at Facebook, I never knew that there was such a rich environment or ecosystem and communities built around mashing together computing and art. I never would’ve thought that — there’s actually a lot when you dig into it.
And so in the same way, there are probably a lot of intersectional communities that are like “Eh, I think it’s crossed with this.” I feel like they do exist. Keep looking for them and if they don’t exist, Find one other person or two other people who want it to exist and create it with them.
#IamthefutureofAI campaign is sponsored by the Ford Foundation and the Omidyar Network. You can watch this and other inspiring career stories on our YouTube channel.
Lia Coleman is an artist, AI researcher, and educator whose work revolves around the interplay of AI technology, art & design, coding, and ethics. Currently, they conduct creative AI research at Carnegie Mellon University.
Lia co-organizes the NeurIPS Workshop on Machine Learning in Creativity and Design, and also works with RunwayML on an explainable ML video series. They were an adjunct professor at the Rhode Island School of Design and created a guide with the Partnership on AI for artists to use ML responsibly.
Their work has been featured in Tribeca Film Festival, Science Gallery Detroit, New York University, the NeurIPS conference, Mozilla Festival, Gray Area, and Wired. Their writing on AI art has been published by Princeton Architectural Press, DISEÑA, and Neocha Magazine.
Lia is an alum of the Massachusetts Institute of Technology (BSc, Computer Science) and the School For Poetic Computation in NYC.