What the Future Holds: Human-Technology Symbiosis
A conversation with Alexander Reben
Is creativity a solely human trait, a thought process constrained by what us humans define it as or is it something that AI will fully master in the near future?
Alexander Reben is an artist and roboticist who explores the relationship between people and machines through his art. This is an important thing to understand as technology gets more complex. Interested in “playing with the human’s role in the loop of machine-creativity” he uses tools such as artificial philosophy, synthetic psychology, perceptual manipulation and technological magic allowing us to experience a future within metaphorical contexts.
From building robots for NASA to filming documentaries entirely through BlabDroids to the world’s first TED talk that was written by an AI and presented by a cyborg he is probing the inherently human nature of the artificial searching for a future where technology will enhance humanity rather than replace it.
What led you to find a creative outlet in AI and how has this influenced your creative universe?
I have used technology in different ways in my artworks, and AI is a natural extension of that creative process. Some of the first work I used AI (a.k.a. machine learning) for was with brain-machine-interfaces in which a system tried to produce artwork that I like based on my brainwave and body-signal response. I was interested in playing with the human’s role in the loop of machine-creativity. Later I used DeepDream and a WaveNet-like algorithm for a video work, Deeply Artificial Trees which is really when I first started using AI in a way that was visible to the viewer. It has expanded my creative toolkit and has provided me with a glimpse into a future of human-machine collaboration. I have no doubt that more tools are on the way which will expand creative and imaginative possibilities.
Where do you find inspiration?
Many places, things I read, see, and hear are certainly a big starting place. New technology, research and news are good resources to keep on top of progress and spark ideas. Sometimes ideas come at random times during conversations, in the shower or from unknown influences. I keep an Evernote notebook of ideas when they come, then I can organize them and come back to some when they have time to develop.
I think this type of intelligence is something that is still so new to the majority of us that when we encounter it we tend to project our humanism, feelings and emotions into it. So in that sense, the art made by an AI generates a bigger reaction to it than maybe a piece made by a man?
I think works which were made by AIs got such big reactions because they were unfamiliar. The public had never seen a generated face or generated text. However, as these technologies become more commonplace, we will see less interest in their inherent novelty, and more emphasis on them as tools of expression. You can see some parallels in photography or film. The first photograph just in itself must have seemed magical, certainly there are accounts of people seeing film for the first time ducking out of the way of a train on screen. However, those initial reactions to the new technology wore off and a camera now is just a tool for creativity. I think the same will happen with AI, and that will lead to new masterpieces.
Once thought of as a solely human trait, creativity is now a debated topic. Do you think that AI can fully master the creative process? Or is a thought process constrained by what us humans define it as?
I think AI will produce output that we perceive as creative (it arguably does that now), certainly with systems that are trained on huge amounts of data (such as GPT-3) seem to be creative in a way. However, that is really a type of “combinatorial” creativity, where old ideas are reassembled into new. This is how many human ideas are formed, so we are already there in a sense. However, I think the real interesting question is if we will ever produce a system which can truly create something which it had not “seen” before in the data it learns from. At that point, we could say the AI would be truly visionary.
What do you think the role of artists, writers and musicians is at the moment?
I think it is the same as it has always been, the tools just change. Mainly to ask questions, convey ideas, inspire, elicit emotion, and produce beauty. In the year of COVID specifically, I think they provide a means of escape in times of confinement and anxiety while giving us a way to expand imagination.
Artificial intelligence is starting to master what it means to be creative but given that there have been so many dystopian stories told of our future with AI, some people might still be skeptical about this technology, can you share with us your vision of the future relationship between humans and robots?
I would distinguish robots and AI (machine learning) as different but related technologies; one is physical while the other is wholly digital. While robots and AI can be combined, they are two different things, you could certainly have a robot without AI and vice versa. I think a lot of what people think about AI comes from popular culture and sci-fi, which tend to deal with AI that is as intelligent or more intelligent than humans. While AI is becoming quite advanced, there is still much uncertainty as to if it will ever be as intelligent as us. More realistically, it can be better than us at specific tasks, say seeing a stop sign in an image or cancer on an X-ray. In these cases, we can see obvious benefits to having accuracy and speed. However, as AI becomes more creative, we may start needing to decide what we want automated and what we do not. For example, AIs are making pretty good music from scratch, certainly good enough to be used for a commercial or to play in an elevator. Imagine a future where an AI makes good music on the fly for your taste – is this as valuable as a song made by a human, does it matter if you enjoy it? It might be more of a philosophical distinction than a practical one. I’d like to see a future of human-technology symbiosis, where we do what we do well, machines do what they do well, and by combining the two we can do better than either would do on their own. Technology to enhance humanity rather than replace it. I explored this idea of machine imagination and collaboration in my recent AI Am I? series where an AI generated descriptions of artworks that were then made in real life.
I’ve spent the year interrogating my needs and rethinking my values: What is for you really necessary and what defines success?
Beyond the basic needs every human has, I think having the physical and mental space for play and imagination is important. Certainly, community is important and having others to discuss and share ideas with is critical. I would measure success on different levels: Do I feel good about and enjoy what I made? Does it communicate well to others? Does it ask important questions? I think my artwork in particular is made for others to experience, so having a venue to show work is critical to its success. This is something that has changed drastically during COVID. The digital has taken over from the physical, and that might last into the future.
At this moment of uncertainty one of the only things we have control over is self-optimization. What have you been doing to find creative solace and emotional relief as well?
Probably tinkering and trying new things with high probabilities for failure. I think continually playing is important, and is something we tend to lose as we grow up. The AI AM I? series was partly an experiment in keeping creative flow by “collaborating” with an AI. Since the AI generated descriptions of artworks in mediums I had never before attempted, it pushed me to try new things and play with new ideas and processes. It helped push me out of the creative funk I think many of us were experiencing.
Based on your Fortune Cookies series, do you think robots have the power to predict the future?
I think anyone who says they have something that can predict the future is trying to sell you something. You can certainly use math, and therefore algorithms, to find the probability of future events occurring. AI might provide ways to increase the accuracy of predictions, but it’s certainly not seeing into the future.
Can you tell us about the TED Talk you created that was written by an AI and presented by a cyborg? How did that work?
“Five Dollars Can Save Planet Earth” was a TED talk generated by an AI and performed on the TEDx San Francisco stage using a robotic mouth. I trained machine learning models on all the TED talk transcripts split into sections, which would for example group introduction text together into one model. I would then generate text from those models and curate outputs that I liked into a presentation. Another algorithm then found images based on the text and I used the PowerPoint auto slide layout to make the slides. It was an investigation into how machine learning could be creative, combined with the absurd idea of having a system use my physical presence to make a presentation on stage. Like much of my work, it leaves space for interpretation and questions, but one main question is: what do we want to automate in the increasingly artificially intelligent future?
What does Utopia mean to you? And does it play a role in your work?
Utopia (besides being unattainable) in a practical sense is balance. It is about the choices we make and the values we hold. What do we want and what don’t we want? Utopia is a direction and not a destination, it is about asking questions which may not have answers. My work looks to the future and uses experimentation, play, imagination and most importantly it uses questions. Making predictions about the future may not be as useful as asking questions of it, because when we make predictions, we tend to surrender ourselves to a set path, while questions leave our possibilities open.