Key Principles for Interpreting Dreams with AI

Kelly Bulkeley

At first glance, the use of AI technologies in dream interpretation offers many benefits. It can provide fast, detailed, and highly personalized feedback about specific dreams. It can apply a variety of analytic perspectives, using both text and imagery to highlight possible meanings. And it can do all this in multiple languages. If the ultimate goal of dream research and education is to give people enough good information to inspire them to explore their own dreams, these are very positive developments.

But should we be careful of what we wish for? What are the possible drawbacks and dangers that come with this new tool? Could the negative effects outweigh the positive uses?

Basic Principles

Common guidelines for sharing and interpreting dreams between people can be applied to AI systems, too. No need for reinventing the ethical wheel here. A dream interpretation system using AI can provide psychological benefits if:

• The dreamer is always in charge of the process and always holds the final responsibility for deciding what the dream means.
• All information entered by the dreamer into the system remains private and can be removed by the dreamer upon request.
• The AI is presented not as a higher authority or divine oracle, but as a source of possible meanings and perspectives.
• The dreamer is encouraged to become involved in real-world communities of dream-sharing.
• The system makes clear it is not a substitute for psychotherapy and if the dreamer feels the need, they should consult a mental health professional.

Such general guidelines are familiar to anyone who has done research or therapy with dreams over the past few decades, and there seems no reason why the same principles should not apply to this new technology, too. Following these guidelines will enable the most positive engagement with AI dream interpretation systems while protecting against the worst abuses. If these basic principles are not programmed into an AI system, that alone should be a cause for concern and wariness.

Towards a Hybrid Approach

AI-generated dream interpretation is not therapy, but it’s not not therapy, either. Current AI systems can already provide people with remarkably insightful interpretations from several different theoretical perspectives. The power of these systems is growing rapidly, going well beyond “dream dictionaries on steroids” (a phrase I used last year and now regret) to approximate the kind of personalized feedback a dreamer could expect from a well-trained psychologist.

This is still not an equivalent to psychotherapy, although it’s easy to imagine therapists and clients using an AI dream interpretation system as a resource in their work together. Future AI-therapist hybrids seem inevitable and potentially very helpful, if the principles mentioned above are built into their functioning.

Natural Therapy

Even bigger implications arise for people who are simply curious about their strange, puzzling dreams. Such people may not be suffering a mental health crisis requiring professional help, but they are still going through psychological experiences that could yield valuable insights, if only they knew how to understand the symbolic language of their dreams. Unfortunately, there are not nearly enough therapists or well-trained psychologists available to help these kinds of dreamers. But a properly designed AI system can address their needs with a speed and ease of use that, over time, will prompt a dramatic boost in their dreaming self-awareness.

Paradoxically, the technology of AI dream interpretation may in this sense qualify as a kind of “natural therapy” as outlined in the final chapter of Robert Kegan’s excellent 1982 book on developmental psychology, The Evolving Self. Kegan offered this term as a way of encouraging greater appreciation for psychologically valuable practices that occur outside a therapist’s office. By “natural therapy” he was referring to “those relations and human contexts which spontaneously support people through the sometimes difficult process of growth and change” (256). In today’s world, nothing is more natural than our daily activities on the internet, using digital tools and interacting with others in virtual spaces. AI systems can provide the kind of simple, thought-provoking dream insights that countless people today are seeking but cannot access through conventional therapeutic resources.