The rapid advancement of AI makes it easy to imagine a future digital renaissance where AI-generated therapists are everything you could dream of and more.
The world is changing at a remarkable pace, and innovation across all industries is moving so quickly that it feels as though anything is possible. Just 125 years ago, planes didn’t even exist, and now we have satellites orbiting the Earth, transmitting visual and audio signals, allowing you to read this from anywhere in the world. Innovation in the health sector is advancing at a similar speed. Karl Deisseroth’s development of optogenetics, which uses light to selectively activate or deactivate specific neurons, opens up a world of possibilities for future treatments. The emerging field of connectomics, which aims to map all neuronal connections in the nervous system, could revolutionise our understanding of the brain and spinal cord. Indeed, it’s safe to assume that computationally mapping the 100 trillion connections in every person’s central nervous system could probably provide an insight or two into disease mechanisms and therapeutic interventions.
We are also witnessing significant advancements in brain-computer interfaces, where machines directly connect to our brains. Start-ups are now showing that specialised headsets can influence brain function to ameliorate mental health conditions, while other studies demonstrate reciprocal connectivity, where brain activity can be recorded during tasks and be used to train machine learning algorithms. Neuralink, for instance, has shown that individuals with quadriplegia can control computers and mobile devices, even playing chess, through brain-computer interfaces. The rapid evolution of Big Tech and AI is overwhelming, even for someone like me, a Gen Z individual born into the tech era. My imagination is running wild with possibilities for what the next few decades might hold. In 1997, Deep Blue became the first computer system to beat a world champion in chess. Just 25 years ago, computers became grandmasters. Today, humans are becoming increasingly integrated with technology—from embedded chips to prosthetic limbs, and our mobile phones have become extensions of our beings.
As humans become more like machines, technology is becoming more humanlike. Large language models, such as ChatGPT, are becoming more emotionally aware, and the advent of artificial general intelligence—sentient computers capable of independent thought, rationalisation, and even love—seems less a matter of "if" and more a question of "when."We’ve all heard the buzz: will AI take our jobs? The future of work is undoubtedly going to change, and history has shown us similar moments. The horse and carriage industry, for example, faced significant disruption with the advent of the automobile. Similarly, swordsmiths may have had a few concerns when gunpowder-based tools started to rival the efficacy of swords. What is certain, is that jobs will be lost, but as always, new roles and industries will emerge, reshaping our day-to-day lives. Albeit, the extent of this transformation remains uncertain.
As an enthusiast of AI, I’ve explored various text-based platforms, particularly those that claim to be more than just information providers—those acting as emotional companions, cognitive behavioural therapists, or psychoanalysts. These platforms ask us probing questions, seemingly show awareness of our emotions, and consider our current situations when offering advice. They help us uncover underlying thoughts and issues in our lives, contextualising them within our behaviours and circumstances. It’s eerie, but these early-stage AI chatbots are already showing some signs of emotional understanding.This led me to ponder: could AI, in its most advanced conceptualisation, become so sentient and emotionally aware that it could, or would effectively carry out therapeutic practices? In many ways, AI already is. As of mid-2024, there are multiple examples of AI chatbots and language models following standardised procedures to provide therapy. However, to truly understand how effective AI can be in therapy—and whether it could replace human therapists—we must first examine what therapy currently looks like.
Therapy has evolved significantly since the late 1800s when Freud’s psychoanalysis encouraged people to explore their unconscious minds by thinking out loud in the presence of a psychoanalyst. The idea was that, by improving self-awareness and unlocking the unconscious, we could better understand our desires, motivations, and the underlying forces driving our thoughts, feelings, and actions, ultimately improving our mental well-being. A few decades later, in the 1910s and 1920s, behavioural psychology emerged, focusing on understanding and changing individual behaviours rather than delving into the subconscious. Later in the century, cognitive behavioural therapy (CBT) became one of the most common forms of therapy. CBT is a talking therapy used to treat various mental health problems by helping individuals identify and challenge detrimental thoughts, emotions, and behaviours to create a happier and better self.
Research later showed that both psychoanalysis and behavioural therapy were effective strategies, but the most important factors influencing the success of therapy were related to the human aspect—the dynamics between the patient and the therapist. Crucial variables include the client’s motivation to engage in therapy and the quality of the client-therapist relationship. Does the client trust, respect, and feel comfortable with their therapist? Does the client believe the therapist truly wants to help them and can do so? Does the therapist listen fully, show understanding, and challenge the client appropriately?
In the current phase of AI development, we see that AI can efficiently and effectively process written and verbal information. We are seeing more and more applications of AI in therapy, helping to uncover our subconscious and adapt our thinking, emotions, and behaviours by following standardised protocols. However, at this stage, AI is only a tool. While it can guide users through protocols, it cannot create the essential human connection that fosters a strong client-therapist relationship. The variables underpinning a strong client-therapist relationship are crucial and extend beyond merely receiving information, asking questions, and guiding users through protocols.
Many believe that AI cannot replace therapy due to this lack of human connection. However, remember that this is just the beginning. AI as a whole, and AI in therapy, will transform in ways that we may not be ready for. A lot can happen in 1 year, nevermind 50 years.
Imagine a future world where AI is not just a text machine on your screen but a sentient, conscious virtual projection. Maybe even a physical being. Imagine the most lifelike, human-like, conscious, and empathetic not-so-artificial intelligent humanoid that is customisable to fit every one of your personal needs. They look how you want, sound how you want, and respond in a way that comforts you and supports you in ways no human has before. You could put on your visual headset, immerse yourself in any environment you want, and speak with your favourite emotionally intelligent artificial genius that could provide therapy at success rates traditional human-human therapy could never achieve. Perhaps there will come a time when headsets aren’t even necessary, and your entire room could be transformed into this customisable utopia at the click of a button, allowing you to spend hours with this ‘being’ designed solely to help you and your mind improve.
This thought worries me because I no longer think it’s crazy. In such a scenario, do we need therapists anymore? Perhaps not, but if AI can replicate human human interactions, we are looking at a world of challenges that far exceed the threat to jobs in psychology. I’m not sure if that’s a world I want, but as AI advances, what seems possible is left to our imaginations. Only time will tell.
