NFTs and the Ethics of AI-generated Art
AI-generated art has taken the social media world by storm. It’s exciting to finally live out fantasies of being a world class artist, bringing to life anything you can think of with impressive results. Even in just the few months that these platforms have granted mainstream access, there have already been huge strides in their ability to come across as genuine art — as if from the hand of one individual and not an amalgamation of millions of different sources.
However, there are already some ethical issues emerging. Questions that are causing some people to pause and consider the implications of AI-generated art. Not just in the sense of how it will impact the livelihood of talented artists, that have spent years or even decades honing their skills, but also questions about IP-rights, transparency and the skewing of data sets towards negative (sometimes violent or otherwise harmful) results.
When it comes to how AI-generated art, or indeed AI-generated content in general, is being used in the world of Web3 and NFTs, there are some more specific ethical concerns that are worth talking about. Revolutionary technologies can change the spaces they operate in so quickly that there often isn’t enough time to really plan for or mitigate some of the negative impacts that they bring with them. But by looking at some of these areas now, we can at least become aware of the issues and consider our own approach to Web3 projects that utilise these technologies.
There’s a vast range of different topics that could be considered here, but to start with here are five issues that AI-generated content brings to Web3 and NFTs and some questions to consider more carefully…
- IP Ownership Rights
The first issue has been discussed for years now — who owns the IP rights to AI-generated art and, related to this, do works breach the rights of artists included (without consent) in the vast data sets used to create generative art?
The first part is about the relationship between the AI platform and the person prompting it. Who is creating the art, the algorithms or the prompter? Who can profit from the result, or claim it as their own? Can an algorithm have ownership of something? It’s a question that is being answered in different ways around the world right now.
The second part of this issue is perhaps the more concerning one, because there are a growing number of artists who are speaking out against having their work included in AI data sets. Is it acceptable that these data sets scrape the internet without consent of original artists? What opt-out mechanisms are in place for those who don’t want their work included?
2. Derivative and Low-effort Collections
This next issue is one that is particularly significant to Web3 and NFT markets. Every time there is a successful collection or idea, there are immediately a wave of derivative projects attempting to ride the attention that collection has gained. AI-generated art makes this process almost trivial for those who want to promote such derivative collections. Do these collections erode the integrity of the space?
They are essentially using AI-generated art to create short-lived, predatory ponzi schemes…is this just the nature of a decentralised sector, or can the wider Web3 community do something about these derivative collections? How much do we value high-effort creative output in the projects we support?
3. False Sense of Security & Fake Community
AI-generated content can also have more nefarious uses, including leading people into a false sense of security that there is more of a team behind a project than there really is. There is a skill to using AI Art effectively, no doubt, but that can also be used to build trust with the intention of exploiting the people that become interested in a project. How can we spot the difference between creative uses of AI-generated content and those used to fill creative gaps or pad out the impression of a project’s depth?
It goes even further than generated art, of course, as entire communities can be generated with bots and fake conversations taking place on Discord to give the impression that there is something real, when it reality it is just a hollow facade. This will become more of an issue as time goes on, as the ability to distinguish between real users and AI-generated ones will become harder. What can we do to help combat this danger? What are the red flags to look out for, or signs of a good community that can’t be faked?
4. Lack of Symbolic Meaning
Beyond the look of a piece of art, lies its meaning. What story does the artwork tell? What are the motivations of the characters in it? The history of the buildings, landscapes or objects depicted? How is the meaning in a piece of art generated?
AI algorithms are great at mimicking the associations of meaning and connections that already exist, but they are not good at creating their own (nor are they really designed, at the moment, to do so). Where innovation of meaning does come into play, it is often through randomisation — a shotgun approach to the creation of meaning, that can be directed by those prompting the output of the algorithms. Are we losing the soul of art through AI-generated platforms, or are we discovering new channels for it to emerge? A question worth contemplating on more deeply.
5. Threat to Artists
There is also an economic and creative issue at play here, which is that AI-generated art is threatening the livelihoods and appreciation of many artists that have spent decades honing skills now generated on the fly. The manner in which brushstrokes are made is mimicked with increasing accuracy; colour and design theory copied without understanding, even though the results can often be pleasing (at least at first glance); and signature styles are being appropriated without concern for their original source.
It is true that all industries must adapt to our new technological age — and artists have always been doing this, as the rise of digital art has shown. With new tools come new possibilities, and perhaps AI-generated art is just that… another tool. But perhaps it also threatens something more fundamental, that lies at the heart of human creativity and our ability to bring our imaginations to life in ways that have weight to them. Does AI art lose some of this emotional gravity? Can we find ways to appreciate and value our artists, while seeing their many hours of work boiled down to a few prompt words? How can artists use AI-art to enhance their own work and still remain ahead of the curve, given their talents?
Conclusion
All of these issues highlight some of the challenges that AI-generated content presents to the current landscape of Web3. However, it’s also clear that it will become a central component of how Web3 evolves and it will be interesting to watch as we go into new areas both positive and negative — many of which we can’t easily conceive of.
One of the most important ways to mitigate against the negative impacts listed above is to develop strong, lasting connections with people within your Web3 networks.
Within the context of communities with a strong sense of purpose and identity, AI-generated content can become a tool that rapidly advances the mission of any community and ability to achieve its goals. When used for fraudulent or low-effort cash grabs, however, AI-generated content only lowers the barrier of entry for bad actors in the space.
This can be countered by a broader evolution of the Web3 community that values authenticity, complex design theory, and delivering on stated goals rather than the hope of hollow roadmaps or successive cash-grab mints.
Once we change our approach to investing and participating in Web3, then the ethical problems of AI-generated content become less impacting and we can focus on the amazing tools that this new field of technology brings to our creative capacity and ability to create meaningful communities. But it’s up to us to make sure this happens, because the tools themselves have no concern how they are used or the impact they have on what we are collectively creating.
