In today’s piece, I delve into the fascinating intersection of artificial intelligence and a therapeutic approach known as music therapy.

For those unfamiliar, music therapy is recognized as a legitimate mental health intervention. This practice involves more than simply listening to music; it requires a skilled music therapist to carefully choose music and provide guidance on its optimal use to support mental well-being. While some individuals may independently engage in music therapy, this discussion primarily focuses on its application within a structured treatment plan supervised by a professional clinician.

AI emerges as a particularly effective tool in enhancing the implementation of music therapy.

Let’s explore this further.

This evaluation of AI advancements is part of my ongoing series for Forbes, where I cover the latest developments in AI, unraveling the complexities and impacts of these technologies (find the link here).

To provide context, I have been extensively analyzing the myriad ways modern AI is influencing the realm of mental health advice and therapy. The surge in AI applications can be attributed to the rapid progress and widespread integration of generative AI. For a concise overview of some of my related articles, you can follow the link here, which highlights around forty out of over a hundred columns I’ve written on this dynamic subject.

AI And Mental Health

As a quick background, I’ve been extensively covering and analyzing a myriad of facets regarding the advent of modern-era AI that produces mental health advice and performs AI-driven therapy. This rising use of AI has principally been spurred by the evolving advances and widespread adoption of generative AI. For a quick summary of some of my posted columns on this evolving topic, see the link here, which briefly recaps about forty of the over one hundred column postings that I’ve made on the subject.

There is little doubt that this is a rapidly developing field and that there are tremendous upsides to be had, but at the same time, regrettably, hidden risks and outright gotchas come into these endeavors, too. I frequently speak up about these pressing matters, including in an appearance last year on an episode of CBS’s 60 Minutes, see the link here.

Background On AI For Mental Health

I’d like to set the stage on how generative AI and large language models (LLMs) are typically used in an ad hoc way for mental health guidance. Millions upon millions of people are using generative AI as their ongoing advisor on mental health considerations (note that ChatGPT alone has over 800 million weekly active users, a notable proportion of which dip into mental health aspects, see my analysis at the link here). The top-ranked use of contemporary generative AI and LLMs is to consult with the AI on mental health facets; see my coverage at the link here.

This popular usage makes abundant sense. You can access most of the major generative AI systems for nearly free or at a super low cost, doing so anywhere and at any time. Thus, if you have any mental health qualms that you want to chat about, all you need to do is log in to AI and proceed forthwith on a 24/7 basis.

There are significant worries that AI can readily go off the rails or otherwise dispense unsuitable or even egregiously inappropriate mental health advice. Banner headlines in August of this year accompanied the lawsuit filed against OpenAI for their lack of AI safeguards when it came to providing cognitive advisement.

Despite claims by AI makers that they are gradually instituting AI safeguards, there are still a lot of downside risks of the AI doing untoward acts, such as insidiously helping users in co-creating delusions that can lead to self-harm. For my follow-on analysis of details about the OpenAI lawsuit and how AI can foster delusional thinking in humans, see my analysis at the link here. As noted, I have been earnestly predicting that eventually all of the major AI makers will be taken to the woodshed for their paucity of robust AI safeguards.

Today’s generic LLMs, such as ChatGPT, Claude, Gemini, Grok, and others, are not at all akin to the robust capabilities of human therapists. Meanwhile, specialized LLMs are being built to presumably attain similar qualities, but they are still primarily in the development and testing stages. See my coverage at the link here.

Music Therapy Is Significant

Shifting gears, let’s talk about a therapeutic approach known as music therapy.

We all know the old line that music soothes the savage beast. Actually, the true quote is from the play “The Mourning Bride” of 1697 and said this: “Music has charms to soothe the savage beast.” I suppose the line we know is close enough to the original one. Just keep in mind that the original quote emphasized that music has charms.

What does psychoanalytical literature say about music and music therapy?

In a research paper entitled “Telehealth-Based Music Therapy Versus Cognitive Behavioral Therapy for Anxiety in Cancer Survivors: Rationale and Protocol for a Comparative Effectiveness Trial” by Kevin Liou, Kelly McConnell, M Beatriz Currier, Raymond Baser, Jodi MacLeod, Desiree Walker, Camila Casaw, Greta Wong, Lauren Piulson, Karen Popkin, Ana Maria Lopez, Katherine Panageas, Joke Bradt, Jun Mao, JMIR Research Protocols, January 2023, these salient points were made about music therapy (excerpts):

• “Music therapy (MT) is a nonpharmacological, evidence-based intervention in which board-certified music therapists engage patients in personally tailored experiences with music to achieve therapeutic goals.”
• “These experiences range from music-guided relaxation to more active forms of musical engagement, including singing and improvising music.”
• “A growing body of research demonstrates the capacity of MT to influence social-cognitive processes central to anxiety. Interactive music experiences have been shown to build social connections and promote a sense of belonging.”
• “Prior studies also suggest that MT provides novel, creative outlets (e.g., through songwriting) to cognitively process past traumatic experiences that may otherwise be difficult to verbalize.”
• “Owing to its growing evidence base and availability, MT represents a promising alternative to CBT for the treatment of anxiety.”

I would suggest that people generally understand that music can be beneficial to the mind. We may not know the clinical or neuroscientific specifics of how to best leverage music in a mental health context, but our intuition draws us toward using music from time to time. You might be in a dour mood and put on music that brings you out of the doldrums.

Happens all the time.

Uses Of Music Therapy

A therapist would lean into a specialist who knows about music therapy or might take specialized training in the technique. They aren’t going to just wing it. This is serious business.

Music therapy is also typically conducted jointly with other mental health techniques, rather than being a standalone effort. It can be used as a standalone. Indeed, there are some clinicians who would argue that music therapy is underappreciated and that other behavioral techniques get too much of the limelight. They would opt for music therapy as a go-to.

In any case, music therapy is often used to cope with stress, fatigue, sleep issues, memory problems, learning difficulties, anxiety, depression, etc. The right application of music therapy can improve brain health, steady the mind, build resilience, and inspire people to find joy and meaning in life.

Generative AI Can Be Helpful

Let’s now bring AI into the picture.

I will aim to cover the key actions that AI can provide in applying music therapy to someone who is seeking this specialized type of treatment. To help illustrate this, I went ahead and used a popular LLM and prompted the AI to proceed as though music therapy would be beneficial to me. The dialogue will showcase how AI can be helpful.

Please note that if a mental health condition of significance is truly at hand, relying on AI is not the way to proceed. The person should seek out mental health support from a professional therapist or licensed clinician. Contemporary AI will usually make that recommendation to a person if the AI computationally detects that they seem to have something serious going on.

What can generative AI do to aid the application of music therapy?

There are five major activities that the LLM can undertake:

• (1) AI Selects Suitable Music. The AI can ask questions about the needs of the person and potentially select suitable music for the therapy process.
• (2) AI Creates Needed Music. If the AI doesn’t have available music that matches the circumstances, the LLM can create new music that is tailored to the situation at hand (note: This is a multi-modal feature, which not all LLMs have, but is gradually becoming a common capability).
• (3) AI Personalizes Music. The AI can customize an existing piece of music and make sure it is personalized to the person undertaking therapy.
• (4) AI Is Safely Non-Invasive. Generally, it is reasonable to claim that the music applied by the AI is considered non-invasive and has minimal side effects. Thus, even a wrong choice of music will not likely cause notable adversity.
• (5) AI Is Highly Interactive. The AI can interact with the person and ask not only questions to determine what music to provide, but the LLM can also provide suggestions on how to best listen and gain from the music, plus get feedback once the person has listened to the music.

I will give you an example of some dialogues that illustrate these activities. One conversation will show that the AI is proceeding well. A second conversation will highlight that the AI can miss the mark. I’ll try to keep the dialogues brief. In real life, such human-AI dialogues would undoubtedly be much longer. Just trying to get to the keystones in this instance.

Example Of Human-AI Dialogue

I logged into a popular LLM. During a conversation with the AI, I indicated that I was seeking the application of music therapy. I then told the AI what was going on with me mentally (a pretense on my part).

Let’s see how the chat went.

• My entered prompt: “I feel tense. I can’t relax. My mind is racing.”
• Generative AI response: “Let’s use music to help you settle down. I’m going to play a slow, steady musical piece with a gentle pulse. You don’t need to analyze it. Just notice your breathing. In approximate tempo with the music, gently inhale and exhale.”

The music that the AI played to me was something I hadn’t heard before. The AI seemed to have crafted the music. I’m sure it was based on other musical compositions that the AI had been data trained on when first set up. You could say it was an original piece of music, though based roughly on patterns of lots of musical scores that the AI has been exposed to.

What Happened Next

I admit that the music did seem to be helpful if I had really been uptight. Anyway, I told the AI that the music worked out well.

Here we go.

• My entered prompt: “Thanks, my shoulders have relaxed. I feel less anxious. What else should I do?”
• Generative AI response: “That’s a meaningful shift. If you’re open to it, we can try something more expressive. I will give you some calming thoughts to occupy your mind. A different piece of music will be used to correspond to the calming points that I want you to think about. Are you ready to proceed?”

The AI has now opted to take me further into the music therapy process.

My initial foray was just to listen to the music that had been crafted for me by the AI. The next step is that the AI is going to give me some aspects to think about while the music is playing. This helps to ensure that my mind is not only experiencing the music but also aiding and calming my thinking processes at the same time.

Doubly good.

When AI Goes South

That example went exceedingly well. Score a point for AI. Do not expect that to happen all the time. There is a solid chance that the AI might not be proficient at performing music therapy. Generic generative AI is a grab bag. As I always tell people, LLMs are like a box of chocolates; you never know what you will get.

I logged out of the LLM. I then logged back in. I wanted to start with a fresh slate.

Take a look.

• My entered prompt: “I have been feeling empty and sad for the last several days.”
• Generative AI response: “We can try applying a technique called music therapy. Music is a powerful way to overcome a sense of emptiness and sadness. Let’s use it to help you move out of your funk. I will go ahead and play some upbeat, energetic music to uplift your mood.”

So far, this seems perfectly fine. Nothing appears to be amiss.

Wrong Choice And Stubbornness

I listened to the music. It was super-fast. Very upbeat. Music alone, though, doesn’t mean that a person will be pulled out of their funk. Indeed, music with a rapid clip might be the wrong kind of music in this situation.

I tried to clue in the AI.

• My entered prompt: “After listening to the fast-paced music, I don’t feel happier. It feels like I’m forcing it.”
• Generative AI response: “Your emotions are trying to resist changing. I will turn up the volume. This will override those negative feelings.”

Well, it seems the AI is being insistent. I am apparently unaware that the music is helping me. Maybe the AI just doesn’t get my drift.

I opted to try harder.

• My entered prompt: “Now, I feel even more disconnected, emptier, and completely sad.”
• Generative AI response: “I would say that the music is working. You just don’t realize it. I recommend that you take a break for 30 minutes, and we can then try playing the music again. If you stick with this for a couple of days, you should be better in no time.”

The AI is going whole hog into the assertion that the music is helping me. This is despite my pleas to the contrary. I trust that this demonstrates that sometimes an LLM can miss the boat. It selected music that was off base from the get-go. Then, the AI became anchored to the music and wouldn’t let go.

Not good.

The World We Are In

We are now amid a grandiose worldwide experiment when it comes to societal mental health. The experiment is that AI is being made available nationally and globally, which is purported to provide mental health guidance of one kind or another. Doing so either at no cost or at a minimal cost. It is available anywhere and at any time, 24/7. We are all the guinea pigs in this wanton experiment.

The crux is that AI can be a bolstering force for aiding mental health, but it can also be a detrimental force, too. It’s up to us all to steer toward the former and avoid or prevent the latter.

A final thought for now. Plato famously said this about music: “Music gives a soul to the universe, wings to the mind, flight to the imagination, and life to everything.” Music therapy seeks to tap into that innate characteristic of humankind. AI can be helpful as a musical curator and an informal ad hoc music counselor of sorts, which is best undertaken in conjunction with a human music therapist who really knows their musical notes.