Have you ever downloaded one of those mental health apps, full of promises about instant calm and unbreakable resilience, only to scroll through endless notifications and feel more drained than you did before?
That is the experience of many other people, too. In our hyper-connected world, the digital marketplace is flooded with a vast array of “snake-oil” tools claiming to heal almost everything.
Meditation timers, mood trackers, virtual therapists, are all vying for a slice of your attention. The result? The exact opposite of mental relief and calm: overwhelm and cynicism.
What if, instead of adding to the noise, artificial intelligence (AI), could cut through it? What if AI could anticipate your stress spikes before they derail your day, or craft coping strategies as unique as your fingerprint? Could this breakthrough technology, often feared as a job-stealer or dystopian overlord, actually become your quiet, call-on-it-when-you-will ally in your search for better mental health?
As a psychologist who has spent decades coaching high-performers through burnout and breakthroughs, I’ve seen, repeatedly, firsthand, how fragile our mental equilibrium can be. Leaders at the top of their game can be haunted by Catch-22 decisions, disturbed by sleepless nights, or suffer the kind of creeping anxiety that even the most resilient struggle to overcome.
Traditional psychotherapy only helps about 1/3 of people. It is ineffective or causes deterioration in around 2/3 of people. Yes, you read correctly; mental health therapies have, at best, a 70% failure rate. Even more worrying, we still don’t know who is going to be in the 1/3 group that are helped and who will be in the 70% not helped, or harmed group.
Even if you could be sure that you were someone in the 30% group, psychotherapy is not always readily accessible. There are long waiting lists. Even if readily available, today, there is still stigma associated with psychotherapy, and vast numbers of people refuse the help they need because of that stigma.
Then there is the practicality of being able to take time out of your busy day to get help, oh, yes, and the need to have funds to pay for it, too. There are many, many barriers preventing people getting help with mental health challenges.
That could be where AI steps in. Almost all of the LLMs, (large language models), to which you have free access, can help you.
Here are some AI prompts that will give you useful results:
- If a person is suffering the following symptoms, what are the likely mental health causes and diagnoses: [insert the symptoms you are experiencing]?
- What are the 15 most effective techniques to prevent or overcome anxiety/depression/phobias/OCD [insert the problem you are experiencing] without using medication?
- What are the 10 most effective techniques that have been successfully used by those who want to overcome [insert the problem you are experiencing] addiction or stop drinking alcohol/taking illicit drugs?
- What are the 15 most effective ways to improve and protect mental health?
Once you have identified the techniques that are known to be most effective, you might want to ask your preferred AI provider which methods are most effective for people with your lifestyle, your schedule, your location, your circumstances.
Yes, you can use AI to deal with problems after they emerge. Wouldn’t it be better to use AI to prevent the problems arising in the first place? Most definitely.
How can you use AI to prevent mental health problems emerging?
You can use AI prompts to give you a list of the factors most likely to damage your physical or mental health. If you give the AI system a briefing on your life circumstances, and ask it to specify, based on those circumstances, what are the factors most likely to put your health at risk, you can obtain very useful information. For instance, it is known that regular sleep disturbance or deprivation is highly predictive of depression.
AI can perform reasoning tasks that humans find challenging. Our brains are great at spotting big patterns, in real time, but rather poor at pattern recognition over longer time frames, and just as poor at spotting the subtle interactions between patterns that could signal trouble ahead.
In contrast, computers, and in this case, specifically AI, is great at spotting long-term and subtle patterns. So, how can you use that capability? If you have access to them, by analysing data from your wearable devices.
Data such as heart rate variability, sleep quantity and quality, even the cadence of your keystrokes on a keyboard, and your walking pattern, can be used to flag up potential mental health dips days in advance.
If you don’t have access to wearable health devices, you can record your heart rate using a watch, measure your blood pressure using a home sphygmomanometer (blood pressure measuring device). You could log your number of sleep hours, your mood on a scale of 0-10 at the same times each day, your body temperature, your weight, etc. Then feed the data in the AI system and ask it to note any changes that could be indicative of looming mental or physical health problems.
Can you use AI to predict mental health problems, and take the preventative measures necessary to remain mentally healthy? Yes.
You can use a health watch and AI, together, to spot emerging problems. For instance, raised heart rate and blood pressure, coupled with sleep disturbance are warning signals that, if unaddressed can lead to a wide range of physical and mental health challenges. If you are aware of such warning signals and act on them, you can head off a mental or physical health crisis.
How? You could enter your health data into your favourite AI system, regularly, and ask it to alert you to the probability of mental or physical health problems based on the data.
Vocal stress is known to predict mental health problems, and indicate their presence. There are several AI systems that can be used to detect the stress and emotion levels in your voice. If you choose to use AI systems to conduct a vocal stress assessment you can have an instant signal that corrective steps may be necessary.
There are other AI systems, which, if you have a text-based or verbal discussion with them, will analyse your communications for indications of actual or emerging mental health problems.
There is now at least one system that uses natural language processing to converse with you. It draws on cognitive behavioural therapy (CBT) principles to detect patterns in your responses. If you are becoming vexed about a work problem, for example, AI might notice escalating negativity and intervene with prompts such as: “It sounds like this situation is weighing heavily on you. On a scale of 1-10, how intense is that feeling right now?”
There is some evidence that such approaches can help some people reduce their anxiety symptoms within just a couple of weeks. How is that possible? Early intervention in negative thinking is known to disrupt the formation of rumination cycles, those vicious downward mental spirals where worry feeds on and reinforces itself, and makes itself progressively worse.
The quality of help we can get from AI depends on our level of honesty. If we give AI honest input, its detached systems, with the right prompts, enable us see what we can easily miss in our fog of stress, despair or fatigue.
For instance, you can prompt or ask AI to interact with you as if it were a world-class psychotherapist. You can ask it to invite you to share your most frequent thoughts on the challenges you face, and then suggest more empowering thoughts and interpretations. You could ask AI for alternative framings of your current perceptions.
One of the many reasons that psychotherapy is so hopelessly ineffective (around a 70% failure rate) is that there are multiple ‘schools of psychotherapy’ which require their followers to use prescribed methods.
The same methods are used on every client, whether those techniques will be effective or not. The false assumption is that those techniques WILL work with everyone; they don’t, they won’t and they can’t.
Even worse, there is no known means of assessing which psychotherapeutic approach will work with any given person. That means, if you are a client who is seeking mental health support, it is a blind search. You don’t know what each psychotherapy school advocates, and even if you find out at your first therapy session, you don’t know whether the method offered is best for you.
AI doesn’t know which approach is best for you either, but it can enable you to go through the trial-and-error process of finding which approach is best for you, without costing you a fortune in trying out different therapists with different approaches.
If you wanted you could prompt AI to ask you questions to identify which of the therapy approaches is best for you, and then use that method.
For instance: “List the most commonly used methods of psychotherapy. Taking each in turn, ask me a series of questions to determine which approach seems most suited to me, my problem, and my circumstances.”
It is known that the more personalised an AI approach to mental health assistance, the more likely it is to succeed. That means it is up to you to give the AI as much useful, honest, information as possible to enable it to personalise its help.
You could start using AI to improve your mental health by asking any of the general LLMs to conduct a search of the most effective AI tools. Once you have that list, you can ask which are the most effective AI tools for your specific mental health challenge.
Over time, you can train your AI app or LLM to become the world’s best expert on your mental health. An expert with perfect recall, an expert which can instantly access almost everything that has ever been published online about your mental health challenge, an expert that can give you the best possible, and most personalised approaches to addressing your challenge.
You might ask, why is an experienced psychotherapist and coach so strongly advocating using AI. Why is that person arguing a position that will surely deprive them of work?
Because, there are, you may find this difficult to believe, many mental health “professionals” who are, frankly, about as much use as a cat flap on a submarine, and whose primary concern is making and keeping you dependant on them.
The more people with mental health problems who use AI, the faster such “professionals will be out of business or rendered redundant.
There is another reason. The more a person takes self-responsibility for their physical and mental health, the healthier they are. Using AI to improve your mental health is an act of self-responsibility.
That is not to say that there some circumstances under which seeking the help of a mental health professional is anything other than the best option. There are many such circumstances, and the more serious the mental health challenge, in my view, the more accessing an ethical and honest professional is the best option.
At the other end of the scale, the very best therapists should be able to help you solve the mildest mental health challenges in just one long session.
Imagine a world where you can find the best methods to deal with your mental health challenges, instantly. Imagine that you can learn about the techniques to build your mental fitness, instantly. With AI, that world already exists.
The challenge for you, should you choose to accept it, is to implement the wisdom that AI has collated for you from the best available online information.
AI can give you the techniques, only you can turn them, first, into better mental health and then into mental fitness.
Professor Nigel MacLennan runs the performance coaching practice PsyPerform.