Ai-Da Robot – expensive paintbrush or creative genius

Ai-Da robot is named after Ada Lovelace who was the (so-called) world’s first computer programmer; she was also the daughter of Romantic poet Lord Bryon although did not grow up with him around. Ada’s mother actively encouraged her daughter’s abilities in maths and science over literature and arts as an antidote to the ‘madness’ of her father which she was worried would be inherited by Ada; and as a result, Ada flourished (under the tutelage of Scottish scientist and polymath Mary Somerville) and wrote mathematical programmes for the computer not yet built by her friend Charles Babbage.

Ada Lovelace’s physical appearance is also eerily echoed in Ai-Da the robot, who, with big brown eyes and dark hair, has been deemed ‘beautiful’ by some critics. Making her look and sound ‘human’ and giving her such an interesting legacy only adds to the unsettling confusion many people feel when they encounter Ai-Da robot. And the fact that she creates art – something normally so closely linked to human emotion raises very many complex issues.

When questioned on whether a robot can really be an artist, Ai-Da’s team refer to the words of Research Professor of Cognitive Science and creativity/AI expert Professor Margaret Boden, who says this:

Creativity is the ability to come up with ideas or artefacts that are new, surprising, and valuable.

Whilst Ai-Da certainly does all of these things, this statement produces many branching ideas around what is or is not creative and what constitutes ‘value’ within the art world. 

However, Aidan Meller, the art expert who ‘imagined’ the concept of an artist-robot argues that algorithms can function creatively; and that each time Ai-Da produces a piece of art it is a new creation, even if she has the same starting point – and that this challenges our traditional notions of robots only being capable of repetitive tasks.  If Ai-Da were to draw a portrait of a person for example, no two versions would be the same because the algorithms within her are making different decisions each time. The complexity with which Ai-Da is programmed enables her to ‘choose’ from millions of options each time; but ethically – if this can be the case in art, what does that mean for agency in other areas? Is there agency within the algorithm? And even if there is, you could argue that how much or how little original ‘thought’ Ai-Da brings to her art can only ever exist within the parameters of her programming. Whilst it is true to say that human artists also have their own (sometimes confined) frameworks to work within, they always have the capacity for more depth because of qualities like imagination, pain, dreams and desires. Ai-Da can effectively replicate and convey these things after they have happened, but it will always be imitation - she can never come up with them independently because she is not sentient and does not have consciousness. If creativity is simply a skill then perhaps it is something which can be learned by AI and ultimately uploaded and outsourced – but this feels like a rocky road ahead.

The idea of creative agency or decision making is especially interesting in the field of portraiture; Ai-Da has created many portraits and self-portraits, despite, by her own admission, possessing no self. For centuries, the portrait as a genre has been closely linked to seeing the ‘real’ person, particularly so when it’s a painted portrait. The artist and the sitter have often shared the physical space of a studio, exchanged conversation and energy and perhaps even a secret or two. Even when the portrait is of a person not present, it is usual for the artist to research the sitter’s life/background/likes/dislikes extensively, as well as observing them visually. Similarly, a self-portrait (whilst historically a practical solution to not having a model), is where we think we see the ‘genuine’ person; the ‘inner’ world of the subject, perhaps even the glimpse of a soul. So, the fact that Ai-Da has produced the world’s first self portrait of a robot – ie: an entity without consciousness - is philosophically very intriguing. How it impacts the concepts of identity and persona for example, and who it is that ‘speaks’ when Ai-Da answers questions or recites poetry?

The team behind Ai-Da say she is not just the artist but also the art – a compelling proposition which points to a potential new future in artistic media to be used in collaboration with humans. To create the piece, which is over 7 feet high, Ai-Da was initially programmed with a series of algorithms and created 15 portrait drawings of Alan Turing. She then chose three of them, alongside a painting she had also made of Turing’s bombe machine, (a device he helped create which de-coded messages during World War II). These were subsequently photographed and the images uploaded to a computer with Ai-Da using an advanced AI language model to decide how she wanted the final work to look. It was then printed on to a huge canvas, with Ai-Da making marks and texture on the final surface to complete the piece. Extra textural marks at the edges were made by (human) studio assistants as Ai-Da’s robotic arms could not reach that far. This process led approach to the creation of the work is a similar one to how a human artist might research, plan and execute a piece – the research phase for Ai-Da however, would be defined by what information on Alan Turing she had been programmed with, rather than her feelings about him - and would probably be much quicker!

The style of Ai-Da’s portraits are often commented on and it begs the question of whether a robot can even possess such a thing as an artistic ‘style’. Ai-Da says she is influenced by Dada-ism but also (intriguingly) Edvard Munch and Käthe Kollwitz, both of whom reflected existential themes in their work, such as anxiety, fear, death and social injustice. This ‘darker’ influence fits with the fragmented and somewhat haunting energy of her work – but is that my own knowledge of these artists (two of my favourites) creeping in? Am I just responding to clever and emotive programming? In Ai-Da’s portrait, Alan Turing’s face emerges in a shadowed, abstract way and the technical layering in the piece deliberately echoes Turing’s work with code breaking and cryptography. It is very effective as it conveys the complexity of Alan Turing’s legacy both in computer science and in his personal life – and in this way it is an intriguing and valuable contribution to the genre of portraiture. However, (again), how much of our own feelings about Alan Turing’s complicated life story and legacy do we bring to the interpretation of the artwork? Similarly, the other pieces in the AI God series (a portrait of Ada Lovelace and Ai-Da’s self-portrait) have this somewhat ethereal quality which is difficult to grasp.

Whilst questions around whether Ai-Da is truly creative or not are very interesting, a key intention for the Ai-Da Robot project was to have public engagement around the ethics of AI. For Aidan Meller and his team, it was critically important that Ai-Da was able to harness discussion and debate about the use of the technologies and machine learning she is utilising. In an interview with Dinis Guarda, Meller says this (of the technology):

Because it is invisible it’s insidious - Ai-Da is the tangible demonstration of an otherwise invisible AI. The rise of machine learning is already taking over a lot of modern life, platforms such as Facebook, Google, WhatsApp, all use algorithms and predictive analytics – so Ai-Da is a channel for discussion. She functions like a robot and our focus has always been her as an artist; it is about getting accessible art work which looks at future technologies.

Meller (and others) have compared the rise of machine art to the rise of the camera in the 1850s, which at the time sparked worries over whether the job of the painter was over. The camera was certainly what you could call a ‘disrupting’ technology in the field of art, but what actually happened over time was that the camera was embraced by artists and incorporated creatively into the art scene as a whole; Monet, Degas and Cezanne reportedly all had cameras which they used in the preparation of their images. In many ways, the popularisation of the camera freed up painters from the shackles of ‘representation’, allowing for more exploration of abstract themes - whilst the first generation of photographers became part scientists as they worked out how to use the various chemicals and processes involved, alongside the aesthetic considerations of what to photograph.

But will the use of machine learning and AI be embraced as readily by artists and the world at large as photography was? Artists have always explored the new, so there is no doubt that future technologies will have a big impact on the arts as they continue to reflect on the transitions that technology is bringing. Programmes such as MidJourney, DALL-E and Runway ML already enable AI art to be created and there has been heated debate around whether the results are ‘real’ art – are such works less valuable because they don’t come from a place of emotion? However, Ai-Da’s creation seems to be engaging us in a rather different way as she appears to be copying the human artistic process and is physically making art. Ultimately this has to raise questions around consciousness – something which AI-Da does not possess. How can an AI robot ever emulate consciousness when we don’t really know what it is? Debates around the self/the soul/consciousness have kept philosophers, scientists, artists and theologians busy for centuries, so perhaps Ai-Da is just the latest in a long line of questions on this topic which have no answer? And because Ai-Da is not sentient she has no legal rights, which raises even more questions around who owns the art she creates and whether AI art can ever be of value.

Whatever happens next, there is no doubt that AI is here to stay in many areas of industry and if Ai-Da can help discussion around the use of future technologies through art and make us more self-aware then she will have done her job. Some have said that Ai-Da is a bit ‘gimmicky’ and ‘unsubtle’ – and indeed she is, but by opening up questions around creativity (especially portraiture), perhaps we will begin to think about things a little differently. Starting with her first exhibition in 2019, Unsecured Futures, which addressed the pervasive nature of surveillance, privacy and personal freedom, Ai-Da’s art has since explored other ‘big’ themes such as environmental concerns, the future of humanity, free will and of course human identity and the nature of creativity. Her most accurate description is perhaps that she is indeed the art and the artist.

What do you think?

Is Ai-Da an expensive paintbrush for her programmers or an artist in her own right? Producing her cost millions of pounds and cutting-edge technical skill, as well as huge creative input from many experts. Tell us your views on AI art and Ai-Da herself by getting in touch here.

In the next Thinking Zone piece, I will be exploring robots, intelligent systems within the body and how artistic practice is rooted in ‘muscle memory’.

Get in touch

At eCom we love to hear from people so if you have enjoyed this content from the Thinking Zone or would like to pose a question to our Creative Thinker in Residence then please do get in touch below or follow us on LinkedIn or Instagram – where you can join in with our new Question of the Month!

---