Algorithmic Visions: Weighing the Benefits and Costs of Psychedelic AI Art

Psychedelic AI art is everywhere – or at least, I’m seeing it everywhere (because many accounts/people I follow on social media are focused on psychedelics). I see psychedelicised AI art used as images for social media posts, blog posts, event listings, websites, and podcasts – in all kinds of online content, really.

I mentioned one negative effect of this trend in a previous article: the images start to look quite homogenous. There is something about seeing psychedelic AI art become so commonplace that doesn’t sit well with me; there are some costs of its prevalence that I think are worth drawing attention to. But I don’t feel wholly pessimistic about the rise of psychedelic AI artwork; there are definitely benefits too, in a way that is specific to psychedelics. Let me first sketch out what I think the advantages of this type of art are.

Increasingly Better Depictions of Psychedelic Experiences

One of the main upsides to psychedelic AI art I’ve noticed is that people can recognise some of their own psychedelic experiences in the images they’re seeing. DMT users have used AI to create images of the entities they’ve met in the DMT space, and these have also been featured in an illustrated guide to these entities, written by David Jay Brown and Sara Phinn Huntley and published by Simon & Schuster. Not everyone will find these depictions relatable, but many do. And for those who get that eerie, flashback-like feeling looking at these images of strange creatures, this can be a useful way to recall what their valuable DMT experiences were like.

This point about using art to recall psychedelic experiences applies to non-DMT-focused AI art, too. People may notice scenes, figures, geometry, patterns, colours, and themes that take them back to their experiences with other kinds of psychedelic compounds. Art can be a powerful tool for integration, which, as I’ve argued elsewhere, depends on our ability to recall these often hard-to-remember experiences. To be taken back to a psychedelic experience, via art, we may be taken back not just to its visual aspects but also its emotional, spiritual, and somatic aspects. All this is to say that psychedelic AI art may be useful in helping people reconnect with the memories of their altered states of consciousness.

What makes psychedelic AI art particularly useful, in this respect, is the fact that people can generate images based on whatever (highly personalised) prompts they give an AI image generator. And the image is created in a matter of seconds. No one has to wait around for a highly skilled artist to pour hours and days into creating a piece of artwork that will resemble some people’s psychedelic experiences (but perhaps not your own). In addition, generative AI may be able to depict aspects of trips in a uniquely accurate way.

The Rise of Psychedelic AI Art Means Aesthetic Losses

I would say there is a cost to overrelying on psychedelic AI art. If we move away from human representations of these experiences and continue the trend of using algorithmic visions, I see a few potential losses here.

One is that some aspects of the psychedelic experience won’t be depicted (or won’t be depicted well): this is related to my earlier point about psychedelic AI art having this homogenous aesthetic. Relatedly, I’ve noticed that AI depictions of psychedelic figures (e.g. Terence McKenna) – like its depictions of anyone real – are often inaccurate. The AI image often resembles the person, but it’s clearly not a faithful representation of them. The image is essentially of a different (imaginary) person. I find that weirdly unsettling.

Secondly, AI art is not human; it lacks the human touch. So when looking at this kind of art, one is not seeing a person’s attempt to recreate or visualise their experience. This can make the AI-generated image less impactful, despite it being technically impressive. Once we realise how the AI image is created, it can lose its meaning and weight. Thirdly, similar to the rise of ChatGPT and similar AI writing tools, if we overrely on AI to create psychedelic art, this can disincentivise people (both hobbyists and professionals) from spending the time and effort creating their own psychedelic artwork. This would be a massive loss (for both artists and the public). We would see a loss of artistic skill, which develops through practice, as well as less interesting and impactful art in the world.

Does Relying on AI Art Conflict With Messages From the Psychedelic Experience?

Many people experience psychedelic AI art as giving life to their personal experiences: a confirmation of what those experiences were like. But I’ve wondered how certain consequences of AI art seem to run up against some common positive messages contained in psychedelic experiences. One is connectedness. Specifically, people often come away from meaningful psychedelic experiences feeling more connected to others. I’m not sure if seeing AI art in psychedelic content helps to foster this feeling. To me, it often feels very alien and detached from human experience.

While barely noticed, and sometimes only weakly experienced, the use of human-created art in content helps readers to better connect to the content the person has created. (I honestly feel no emotional reaction when reading AI-generated articles featuring AI-generated images.) Perhaps in the future, AI art can so perfectly replicate human art and photography that it will create an equal emotional reaction (this isn’t something to celebrate, necessarily). But until then, relying on AI art in psychedelic content may lead people to feel more detached from the human voice behind the content.

Another way in which psychedelic AI art can disconnect us is through the loss of employment for artists and the loss of artwork. If we continue to rely on AI art for content, then we will collaborate with each other less in creative contexts. Creativity is essential to our humanity; it connects us in the world of work, and without it, we become increasingly disconnected. Instead, we connect on matters less human and aesthetic: efficiency, productivity, and profit. And related to the earlier point about how human art evokes our emotions, if we are exposed to less human art in the world, this could further disconnect us from the lives of others. That’s something we don’t need, given how much of our lives are lived online already.

One more tangible effect of overrelying on psychedelic AI art, which I see as conflicting with a common message of the psychedelic experience, is the environmental impact of generative AI. As Adam Zewe explains in an article for MIT News:

The computational power required to train generative AI models that often have billions of parameters, such as OpenAI’s GPT-4, can demand a staggering amount of electricity, which leads to increased carbon dioxide emissions and pressures on the electric grid.

Furthermore, deploying these models in real-world applications, enabling millions to use generative AI in their daily lives, and then fine-tuning the models to improve their performance draws large amounts of energy long after a model has been developed.

Beyond electricity demands, a great deal of water is needed to cool the hardware used for training, deploying, and fine-tuning generative AI models, which can strain municipal water supplies and disrupt local ecosystems. The increasing number of generative AI applications has also spurred demand for high-performance computing hardware, adding indirect environmental impacts from its manufacture and transport.

Compare this negative consequence with a common message and attitude contained within psychedelic experiences: concern for the environment. Research also supports this link between psychedelics and increased connection to nature. If those in the psychedelic field want to protect the environment in a meaningful way, then the environmental costs of generative AI have to be considered when creating content related to psychedelics. There are some barriers to making this happen, however. Noman Bashir, who is a Computing and Climate Impact Fellow at MIT Climate and Sustainability Consortium (MCSC), notes in the MIT News article, “an everyday user doesn’t think too much about that. The ease-of-use of generative AI interfaces and the lack of information about the environmental impacts of my actions means that, as a user, I don’t have much incentive to cut back on my use of generative AI.”

A similar problem arises with platforms like Amazon, Shein, and Temu. People often aren’t aware of just how unethical the business practices behind these companies are – and, even when people have this awareness, the convenience of using them often defeats the desire to boycott their use. Moreover, as I discuss in another blog post, increased concern for the environment after psychedelic use doesn’t always translate into pro-environmental behaviour, for various reasons.

Nevertheless, the existing disconnect between awareness and action doesn’t mean the disconnect is inevitable. I think there can come a point at which awareness of a problem reaches a certain threshold, after which the impact of the problem just feels heavier – more real. At the very least, the use of psychedelic AI art should be thought of as an ethical issue as much as an aesthetic one. By doing so, we can better weigh up the benefits and costs of using this form of generative AI.