March 2020 Newsletter
Welcome to a monthly newsletter from Artists + Machine Intelligence
Hello, and welcome to March, and a monthly newsletter from Artists + Machine Intelligence. In the past year, we’ve experimented with formats to keep you, dear readers, informed of program news. We’re restarting this monthly newsletter to share what we’re up to. Let us know what you’d like to see more or less of in the comments below.
Last month, I had the pleasure of speaking at the Met on ways we might advance art-making through AI and open access. Archives, personal or otherwise, are a foundational material in art practices leveraging machine learning, and I was excited to participate, if only to say that AI technologies can be one part (and one part only) of a broader, technological stack, and that GANs have aesthetics, too. Selfishly, the event also reminded me of how I ended up in this field, so I thought I’d share the story with you.
It was 2016, and I was in Paris, attending an arts and machine learning conference at Google Arts & Culture. At the time, I was a filmmaker with the Museum of Modern Art. I knew very little of machine learning, but I knew of DeepDream (including, yes, this Starry Night-inspired rendition). A number of works were installed at the Lab, among them, “ X Degrees of Separation,” by Mario Klingemann and Simon Doury. Positioned as an AI experiment, “X Degrees” claims to “chart the visual connections between works as disparate as a 4,000 year old oil painting and Vincent van Gogh’s Starry Night” (Google’s words) by displaying a chain of artworks (or visual journey) between any two artifacts.
Was the experience playful and surprising? Yes. (It still is.) Are visual sorting tools that let us navigate large collections helpful? Sometimes. Of course, I couldn’t help but think, what happens when visual similarity is no more than a faux amis, the French term for words in a foreign language bearing a deceptive resemblance to words in one’s own language? And better yet, what about moments when visual difference masks underlying similarity?
I also saw works like Ross Goodwin’s Word.camera (including a generative poem in response to a Rothko painting that is, sadly, no longer online — the poem, not the painting), and I fell in love. More of this, I thought. Works that bring our attention back to the primary subject, be it object or landscape, and help us interrogate it, as only humans might.
This is why, in 2018, I accepted an offer to to co-lead Google’s Artists + Machine Intelligence program with my colleague, Kenric McDowell. Now, as we kick off a fourth year of AMI programming, I wanted to share what’s top of mind for me in a more open format (this newsletter), starting with a few resolutions for 2020.
No more ‘AI art’ conferences. Let’s get specific. Picture-making with GANs. Ethics in co-writing. We won’t be able to build a shared vocabulary until we create time and place for specificity. More ‘art with machine intelligence.’ An artist’s practice may be defined by their use of a certain technology, but we don’t call work by contemporary photographers (ahem, Lucas Blalock) ‘photoshop art’, do we? Why, then, do so many writers, curators, and critics insist on the term, ‘AI art’? A longer historical lens, please. As museums shift to rehang collections outside of historical time frames and media, why do curators still revel in art + technology shows? If Picasso’s “Demoiselles” can hang next to Faith Ringgold “American People Series #20: Die” (1967), why not hang Adam Basanta’s flatbed scans with color field painters like Barnett Newman or Mark Rothko, or screen Anna Ridler’s silent film animations alongside Bill Morrison’s decayed film works?
I hope this monthly letter encourages rigorous discussion and enables some sense-making in this rapidly emerging field. Send me an email or a tweet with what you’re thinking about — and ask me your questions about art, technology, and machine creativity. I’ll answer one per week.
Eva
Here’s what else is happening:
We announced six artist grant recipients for our inaugural grant program, produced in collaboration with Google AI and Google Arts & Culture. We are thrilled to be supporting a portfolio of projects spanning generative and interactive film, poetry, spoken word, and sound.
We continued to award research grants, as part of an ongoing collaboration with Google AI to support researchers at universities pursuing creative AI tool-making and research.
To celebrate our collaborators at Google and beyond who make endeavors like this possible, we are introducing a new Q.&A. series. This week’s post spotlights Holly Grimm, an artist and creative technologist at Google, who ‘taps’ Mark Dion as an artist not working with machine learning, but who very well could be. Mark is known for appropriating scientific methods of collecting, ordering, and exhibiting objects, exploring “what gets to stand for nature at a particular time for a distinct group of people” (his words, as written in one of my favorite exhibition catalogues, “The Museum as Muse: Artists Reflect.”) Which artist would you nominate to explore AI techniques?
Finally, as part of CultureHub’s Re-fest 2020, join artist Allison Parrish for a conversation on how style is implicitly or explicitly built into generative systems. With Maxwell Neely-Cohen, Blair Simmons, and Katy Ilonka Gero. (Thursday, March 12, 7:15 p.m., RSVP).
Eva Kozanecka is a content strategist at Google AI, where she explores how diverse media can improve the representation and communication of emerging technologies. She also co-leads Artists + Machine Intelligence, a program at Google that invites artists to work with engineers and researchers together in the design of intelligent systems.
Questions? Feedback? Tweet us or email: ekozanecka@google.com
