Cass Fino-Radin has spent their career on the side of the acquisition nobody photographs: the decade after.

They were in MoMA's conservation department when it started collecting video games, working alongside Paul Galloway on what it means for the Museum of Modern Art to own Space Invaders. Before that, Rhizome. After that, nine years running the private conservation lab Small Data Industries. They are now Vice President of Art and Technology at Canyon, the non-collecting venue for durational art opening in New York in 2027, and Director of the Canyon Media Art Conservation Center, a nonprofit lab for hire that works on other institutions' collections.

We talked about what acquiring a playable work actually commits an institution to, what it takes to put seventy artists' worth of game-based work into a gallery, and how conservation changes now that a coding tool can do in a week what used to take years.

Jamin: You were at MoMA when it started collecting games. What was the work, actually?

Cass: The bulk of the hard work in those years was on the legal side. MoMA is a large institution. It is not the kind of place where you collect a commercial copy of something and show it based on fair use, especially if you want to emulate it. So it really required [Senior Collection Specialist at MoMA] Paul Galloway to track people down. Just managing to find these people in the first place, then building trust with them, then getting them to understand why you are reaching out at all. And once you have that relationship, it is carving out an agreement for the right to exhibit the work, which with games gets complicated because you are selling a ticket for admission. And then for conservation, carving out the right to do things like emulation, or if there is source code to be had, to tinker with it in the future.

Jamin: You went from there into private practice, working with collectors. What did that teach you about the economics?

Cass: Collecting media art is kind of insane. If you are a private collector, it is a depreciating asset in a way. It is not a Basquiat that I collect and flip in ten years. The secondary market is very soft, and it costs money to even just show it or keep it alive or maintain it, always, significantly, forever into infinity. So it tends to attract people who are in it for the right reasons.

Jamin: Canyon opens with World Building, a group show with a lot of game-based work in it. What does it take to mount that?

Cass: On the operational and AV side, it is kind of hilarious to me. We built this gorgeous infrastructure, over five thousand square feet of direct-view, high-resolution LED connected to a Pixera media server, where we can basically paint with video. We can put any file anywhere, play the artist's preservation files, no transcoding down to something compressed. And then we said, how about we do a big sprawling group show with almost seventy artists. And yes, they are all video games, so we have to buy a ton of rigs.

It is probably the hardest exhibition we will ever do for quite some time, and not even on the technical side. The logistics. Our exhibition producer, our curator at large, and our registrar are wrangling so many artists, agreements, and photos. The catalog is going to be a tome.

But it takes the whole premise of what we are hoping to do at Canyon and hands us the most challenging medium to do it with. It is a bit like putting net art in a museum. That work was conceived assuming the viewer is at home, maybe in bed on their laptop, seen across five thousand different hardware setups. Games have a similar problem. You cannot put Disco Elysium in a gallery, let people play with it, and expect them to walk out saying they get it now. These are things that can take seventy hours to reach the moment where you understand.

Jamin: For an artist who wants their work collected, what has to come with it?

Cass: Whatever enters the collection has to be as complete a package as possible. Source code, project files, all of it. If you don't have that, you are limited to emulation, which is a silver bullet for some things and not for many. There might be one particular weird little quirk about that piece of software where it is not going to work in an emulator, and you really need the source.

That is a complicated one for some artists to swallow. If they are talking to MoMA, that is one thing. They have a team of media conservators and an incredible lab designed for this practice, and they can build the trust with the artist where the artist understands why they need to hand this over. But that is very few institutions. Very few institutions have conservators at all, let alone media art conservators.

Jamin: Does any of that change with what has happened to software in the last two years?

Cass: It is hard for me to say much more right now because of everything that has happened with AI and code. For somebody skilled with software, a developer can do things now that would have taken them years.

Case in point. The Cooper Hewitt had an iPad app they collected called Planetary, and we did a case study on it years and years ago in my lab. It was completely obsolete. It turns out the iPad they had put in storage actually had nothing on it. So the only way to show it would be to get the source code up to date and push it to a contemporary iPad, which meant untangling a spaghetti mess of almost a decade of iOS updates, one version at a time.

What happened was a freak case. Because of COVID there was a fan of the app in Australia, a developer with free time on his hands, and he just took it on and fixed it. That never happens. Open sourcing code is not a preservation solution alone.

Now a developer who knows what they are doing could have fixed that in a week.

Jamin: In conservation you have to document what you did. How do you document that?

Cass: Right, and when you are working with a stochastic parrot that is a little hard to do, because if you ask it what it did, it could lie. So the methodology our conservation technician landed on was to have it write Python scripts for what it wants to do, and then you run the script. Instead of the tool going in and doing the thing itself, it is building a tool to do that thing, and that tool is real source code you can read. This is what I am going to do, and this is what is being done, exactly. Code is law. And it worked.

This interview is edited for content, grammar, and style.