Joan Heemskerk on Quantum Web4
Joan Heemskerk on Quantum Web4
Joan Heemskerk of the renowned net.art duo JODI undertook a two-month residency between CERN—the birthplace of the World Wide Web—and Copenhagen Contemporary. The resulting work will be part of a show at Copenhagen Contemporary in 2025. Heemskerk spoke to Peter Bauman (Monk Antony) about the residency and its impact on her practice. They also discuss her interest in quantum computing, Web4 and the continued evolution of the web.
Peter Bauman: It's fitting as a preeminent net artist that you have a residency at the birthplace of the web. Can you tell us more about your experience there?
Joan Heemskerk: I was initially very interested in the fact that Tim Berners-Lee founded the web at CERN. I was captivated by the question: If everything in society now has its binary output, produced by different programming languages, will anyone be able to reverse engineer this in the future? Will they be able to decipher the meaning of the string "Hello, world!"? Can we find out a hundred years from now what the binary messages of today are saying? It was partly speculative.
With a quantum network and quantum computing, there's no need for binary anymore. At the moment, our universal language is binary but that will change. How do you explain binary to a non-human being or someone who doesn't know how to read it? They were already thinking about this when they made the content plate and a record for the NASA Pioneer and Voyager missions.
Peter Bauman: It's such an inspiring place. It seems like you would have enviable access to first-hand research material on the earliest days of the web that you couldn't get anywhere else.
Joan Heemskerk: Tim Berners-Lee proposed a universal language but I discovered he was just fed up with his job. He was printing papers and sending them to the offices next door or he had to use the fax machine. So by creating network access, it was really nice that people could just log into a computer and read the papers. It was an efficiency thing. He didn't want to work that much. But it took three years before he finally got the agreement to make the World Wide Web. When CERN got the NeXT computer, he got the green light to test it with his proposal. But that's all history. It’s almost like a folk story. The legacy of Tim Berners-Lee is more important than any archival material because they only have a couple A4s and the old computer. It was a side project. It had nothing to do with finding particles there.
The particle research at CERN is really bumping particles in the collider and scraping the data of all the collisions. I saw a lot of control centers—really nice! I saw a lot of computer screens. And I got super distracted by all the event displays they have.
One interesting thing I found out in the data center at CERN is that instead of the web, they use the grid network, which is another parallel network. It's a cloud system between data centers and science institutes. They have a separate network.
Peter Bauman: Did you use machine learning or bespoke neural networks at all in your workflow?
Joan Heemskerk: For the video NO-MATTER, I used image recognition. It's funny because I used Hough transform, the same image recognition they used for the particle analyses in the very first bubble chambers. Nowadays, they use that same algorithm for self-driving cars to find the sides of the street. And I used it to create the lines between particle dots.
Before CERN, my project YOLO.360.VR also investigated recognition algorithms. I was in a scrap yard using AI to recognize cars but because these cars were totally crumpled, sometimes it categorized them as a backpack or a bike. It's about the probability of what you see.
AI software is trained so that if there’s an 80% chance you are a human, it puts a square on you and labels it human. But if you lower the probabilities to 20%, then it would start considering even abstract shapes that looked like a face to be human.
Peter Bauman: That hints at the overlap between AI and quantum computing. Can you talk more about that?
Joan Heemskerk: That's a big overlap. In quantum computing, non-binary qubits are crunching data into binary. You could say it's all AI because it works with probabilities. It's not on and off anymore.
It's not black and white. There are shades of gray. If it's dark gray, AI would say it's black. If it's light gray, it says it's white. Then, if it's a middle gray, it becomes an uncertain area.
Peter Bauman: You’ve spoken before of Web4 as being quantum and non-binary-based.
Joan Heemskerk: Yeah, that's my little joke when Web3 came out.
I think of it as one web. It's already an insult that you have Web2 and then Web3. I thought, “Maybe I should make Web4” [laughs].
Peter Bauman: What would be a better way to delineate those transformative changes or are they just not as pivotal as some make them out to be?
Joan Heemskerk: I don’t think they’re as pivotal. People still use and make websites so why change the name?
Peter Bauman: Can you tell us more about what you tongue-in-cheek refer to as Web4? But really you think this is the next—and also first—truly paradigmatic step in the web's evolution.
Joan Heemskerk: We're at a moment in time that’s actually a bit similar to the beginning of the '90s in a geopolitical, security sense. The early ‘90s was the beginning of the World Wide Web, which was indeed invented by Tim Berners-Lee. There was also ARPANET. But it was basically the web for governments, militaries and science or education institutions. It wasn't a commercial web—other than maybe some hackers around.
With quantum networks, which are still in their early baby shoes, it's the same. You have the scientists plus military, government and educational interests. It's at the phase where it's still very local. It feels like dialup with a local modem.
There are three different quantum fields of study: quantum sensing, quantum networks and quantum computing. All three are different but there are some overlaps. When I started working with quantum in 2019, I was interested in the quantum network, but I didn't get the green light, so I ended up making work with quantum computers. That quantum computer could only spit out errors. So I thought, “Let's make something about that then, at least.”
Quantum networks are shrouded in secrecy, especially at that time. Maybe they're a bit more open now, but still not very. It's dealing with secrets because it's about making a safe network.
And it will definitely happen. There are already very small quantum networks. It's already in a test phase where we can make a network with fiber optics and satellites. There’s a lot of industry involved; if they want to scale up, they have to make better repeaters, for example.
In a sense, it's already here. And that fascinates me.
Peter Bauman: Would you put where quantum networks are today, similar to where the web was in the early '90s?
Joan Heemskerk: I would say '80s. But then you see how fast it can go.
Peter Bauman: Especially now compared to then.
Joan Heemskerk: If there's a proposal, then it could go really fast.
Peter Bauman: Are governments and private industry working together on this or is there more of a sense of competition?
Joan Heemskerk: They're working together and not. IBM, for example, is investing a lot in quantum computing. So it's also a question of corporate secrecy. At the moment, quantum computers are these big refrigerator-size objects. They have this parallel to the mainframe. You can access them through a laptop but the machines themselves are huge.
The quantum network is a completely different thing. That could be a table set up or connecting some cities through fiber optics. So how do you transfer information in a quantum network? Basically, you don't transfer it at all. That's the funny thing. You don't transfer anything anymore. There was this analogy with the telephone that you could make a vibration in the air and really send a message further. But in the beginning, there was this plan to make a mouth on both sides where you control muscles with a valve when you speak. Then you only send the muscle information without transferring the air or vibration. You don't transfer the message. That’s what a quantum network is like.
Peter Bauman: That’s a very helpful analogy. You mentioned refrigerator sizes and the connection to mainframes; is that a good way to think about where we’re at in the development of quantum computers? Mainframes from which decade though? Would that put us in the 1950s, 60s, 70s?
Joan Heemskerk: Even the '50s. When the first OXO game with AI was made, it was made on the mainframe. There weren't even CRTs at the time. You had to buy your time and you had to print out the results. It’s the same today with quantum computing but now you access it via a classical computer and through the cloud. But somewhere in a room there's one or two quantum computers.
Peter Bauman: How many quantum computers are there in the world?
Joan Heemskerk: I can make a rough estimate. I would say that every country might have ten or so but some have more, like the States. So I have no idea. It's also changing really fast.
Peter Bauman: So back of the envelope, there's probably between one and a few thousand in the world? How many people do you think are intervening with them creatively? Artists like Pindar Van Arman, David Young and Entangled Others have explored quantum computing but I'm not aware of more than a handful. Again, the comparison to early computers is striking. Who else besides you is exploring them?
Joan Heemskerk: There are a lot of musicians actually playing with quantum computers.
Peter Bauman: That’s another remarkable similarity to early computing, when Max Mathews at Bell Labs made computer music years before the first visual computer art. It was the same with early neural network experimentation, too. When intervening with new technologies, music seems to precede the visual because of its basis in rules. What about visual creative exploration with quantum computers?
Joan Heemskerk: It's quite a new field. That's true. Someone who comes to mind is Libby Heaney. I don't know if she really uses a quantum computer. I myself am not always using a quantum computer. Sometimes you can make artworks with entangled storytelling. It depends on the piece.
Peter Bauman: Some of your work you would consider yourself "really" using a quantum computer?
Joan Heemskerk: I used quantum computers in Q&Q (Qliza) at TU Delft/Qutech with the Quantum Thinker group and the IBM quantum public machine access.
CERN doesn’t have quantum computers, although there are a small number of theoretical physicists as a quantum think tank. During the second half of my residency, I visited the quantum network and quantum sensing labs at the Niels Bohr Institute. Seeing all the optics, lasers and event displays was the inspiration for the work entangled binary network(Hello, world!).
Peter Bauman: What is so fascinating is that it's not an emerging technology that many people are talking about. Other technologies certainly get more buzz but this could be just as disruptive as anything else that's coming.
Joan Heemskerk: Sure. Next year is “quantum year” and some art institutions are looking for artists working with quantum. They ask me, “Do you know other people?” And I have to say, “Sorry.”
Peter Bauman: You've been at the forefront of both of these developments now, thirty years apart. Do you still see yourself exploring emerging technology—like you did—in terms of subverting its conventions? How would you do that in a technology that doesn't have any conventions yet?
Joan Heemskerk: That's not so difficult. It's not always that I'm subverting. That's the thing. It's more like peeling an orange sometimes to find out what it's about. Then I make a blunt interpretation. I think that's it.
In a sense, I still feel like a parasite of technology, surfing the waves.
Peter Bauman: We've touched on your practice and how you've been intervening with these technologies in different parts of your career. How do you think that humanity’s relationship with technology has evolved since you began your work in the mid-'90s?
Joan Heemskerk: I think it changed a lot in the sense that in the '90s, if we talk about pre-net art, I did photography and some CD-ROMs and shit like that [laughs]. I really had to go to a lab because I couldn't afford a computer. Even the first time I made a website, I had to ask someone else, “Can I rent computer time?” It's a bit like now on a quantum computer. It's free to use but you have to ask.
Then the internet came and—beyond just net.art—enabled a global community of artists meeting each other online and offline. I think the community aspect is really important. I see it in the NFT world; I think the major plus is a community of artists. Things speed up when there is a creative community.
But in terms of humans' relationship to digital technology, how humans react is completely different from artists. Artists look at society and see what's happening. If there’s a new steam engine, Turner makes a painting of it. When the industrial age came, Picabia made a drawing about it. Duchamp makes a joke about the coffee machine. It's the same.
But humans, in general, just have to adapt and take the path of least resistance, unless you're a luddite. I think in society, we cannot live anymore without digital. That's a vulnerable situation because we're completely dependent on it. If there is some interruption in the network, you cannot fly, you cannot take the train, you cannot get your money, you cannot drive your car.
Peter Bauman: It’s interesting that you bring up security. How will the arrival of quantum computing impact our current security infrastructure?
Joan Heemskerk: Maybe quantum computers in combination with the quantum network are the solution. Then you don't need encryption anymore; it's already entangled. As I explained before, you don't actually send a message so in that sense there's no threat. Having said that, quantum computers could be a threat to today’s Alice-to-Bob encryption. It takes a lot of computer power nowadays to break that encryption. But a quantum computer could do that in a fraction of the time. It's really good at crunching data. But at the moment, there are too many errors.
Peter Bauman: So quantum computing itself would be safer, but anything that we think of today as encrypted and safe would no longer be encrypted and safe in the age of quantum computing?
Joan Heemskerk: That's why people are already starting to encrypt in so-called quantum-proof encryption that wouldn’t be so easy for a quantum computer to break.
Peter Bauman: What are the implications for the way blockchains cryptographically operate?
Joan Heemskerk: I’m not a tech developer who knows how blockchains or quantum computers are developing. You see the concern if Vitalik made a proposal to hard fork the Ethereum blockchain to adapt to future vulnerabilities. As an artist, I can only make a jukebox of playing endless keys of hacked wallets, like my project, DoAW. In a way, it’s about the same decryption mechanism as the encrypted blockchain.
For DoAW, somehow I got my hands on an algorithm that makes seed phrases. It lets you see the wallet and the private keys. It endlessly shows seed phrases because all wallets that will ever exist are there already. It's a bit like getting a bank account that already has your password. So it becomes a somewhat vulnerable situation. If I can run that software, anybody can run that software. At the moment, bots do it all the time. They look if there are private keys hidden somewhere in GitHub and they grab them.
It's secure and it's not secure—an idea the project explores. It makes music out of these private keys. But the concept behind it is, “Okay, put your cards on the table.” It's like picking up a grain of sand in the desert and saying, “Here is your wallet and it's hacked.” “And this is the other one and I hacked it.” Do I show the vulnerability of the desert? No, I show the vulnerability of the grain of sand.
Peter Bauman: It at least seems like there’s time to prepare. In terms of interventions with emerging technologies like quantum computing, we've seen with net art and early computers that people are going to look back closely at who was doing this at the beginning. Why’s it important for artists to explore these technologies while they’re still in their infancy?
Joan Heemskerk: Tech somehow needs to have artists to play with it. In the early days of net.art, Mozilla—which along with Netscape was one of the first browsers—looked at our website just as beta testers.
Often, artists are also the people who explore technology and say to society, “Hey, you can use it. It’s okay.” It's like in cities: if there's a cheap neighborhood, artists move in and it gets gentrified. It’s similar with technology. It's artists who jump in there and then somehow it gets mainstream.
In the case of the World Wide Web, it became a consumer network instead of a creative maker's network. There are only very few people who make content.
Peter Bauman: With these advancements in quantum computing seemingly on the horizon, why aren’t more people talking about it?
Joan Heemskerk: Nobody talked about net.art in the '90s. I mean, you could count them on two hands. It is on some people's radar, I think.
Peter Bauman: It’s thrilling to think we are on the cusp of a new art form in addition to a new network and computing generation.
Joan Heemskerk: Sure, it might be a complete disaster after all. If they can't scale up, then it will be a case where humanity did this experiment but we never managed.
Peter Bauman: What does the time scale look like in your opinion for when this will all be more widely accessible?
Joan Heemskerk: It could be five, could be ten, could be a hundred years from now. I have no idea. It's going fast, though. Five years ago, you would be happy if there were one or two stable qubits and now it's easy to scale up to over a hundred.
Peter Bauman: It seems like it could be a similar situation to AI, where it caught people off guard when it finally did arrive. AI was always “five years away," but then around 2023, with LLMs and LDMs, it was suddenly here now. Based on what you’re saying, it seems like we might experience something similar with quantum computing.
Joan Heemskerk: The next generation of chips will have AI and neural networks already baked in. They will be here, probably by 2027. That's only in two or three years. Now if you take a picture of someone, the software will recognize that it’s a person. But if you put that software directly in the hardware—in the chip, so you have a neural network in the chip—it will recognize the human immediately. It would already be able to crunch the data before you store the picture. It's a different sensor chip because we’re hitting a ceiling with Moore's law. It is getting to the point where the chips are so small that the only way you can improve the power is by making smarter technology and incorporating AI. And it's scary. It could be completely biased. But that's another interview.
---
Joan Heemskerk is a contemporary artist whose practice examines the Internet and its potential, including cryptography and quantum non-binary computing. Heemskerk, part of the artist duo JODI, pioneered net.art in the mid-1990s by exploring how the Internet's underlying systems shape the relationship between technology and its users.
Their work has been exhibited at MoMA, Stedelijk Museum Amsterdam, ZKM Karlsruhe, ICC, Guggenheim, Centre Pompidou, Eyebeam, Rhizome and many others.
Peter Bauman (Monk Antony) is Le Random's Editor-in-Chief.