Do Humans Make Bad Therapists?

In the first part of this series we looked at some of the advantages of two radically different approaches to therapy: traditional face-to-face therapy, and the emerging field of computerized therapy. We came to the conclusion that people have that ever-important "human touch", while computerized therapy has the potential to reach an audience unwilling or unable to attend live sessions.

In this post we'll take a look at some of the shortcomings of the two forms of therapy.

Apathetic software

There has been a lot of discussion about the drawbacks of computer-based therapy in academic circles. In brief, here are some of the most common objections:

• As I mentioned in my post on the use of artificial intelligence in therapy, truly intelligent robot therapists still belong to the realms of science fiction. It is possible to create the illusion of high-level functions such as understanding, flexibility and empathy (and there is evidence that this can be successfully incorporated into computerized therapy systems). However, this simulation can only be taken so far: under the hood, a computer is a slave to its algorithms, and doesn't "care" about the patient or their outcome.
• How should automated therapy systems be regulated? In many countries "Clinical Psychologist" is a protected title, and there is a professional organization mandated with ensuring proper training and adherence to ethical standards. On the other hand, anyone can put a therapy product on the internet and make unsubstantiated claims about it.
• There are security concerns for systems that are available online. The worst case scenario is hackers accessing highly sensitive clinical data. Technology exists to make breaches difficult and unlikely, but there is no such thing as a perfectly secure system.

The pitfalls of being a human therapist

People often assume that humans are exceptionally well-suited for being therapists, and the ultimate goal of computerized systems is to aspire to the same lofty peaks of therapeutic excellence. This is fundamentally misguided, as there are some disadvantages of human therapy that should not be overlooked.

• The first problem is inconsistency and variability among mental health professionals. Let's be honest with ourselves: some therapists have not been properly trained, and others are downright incompetent. Furthermore, even the best therapists have off days due to sickness, tiredness, boredom, or personal issues. Just because a client is seeing a human, he or she is not assured a high standard of care.
• Therapist drift is a phenomenon whereby a therapist, either intentionally or unintentionally, moves away from the core principles of an evidence-based technique over time. This is problematic as the resulting treatment is typically less effective. Similarly, research studies are leading to new findings all the time, and sometimes overturning previous results. Disseminating new information to practicing clinicians is a significant challenge, and there are many therapists in the field unknowingly using outdated techniques.
• Humans are not nearly as rational as we like to think we are. Over the past few decades investigations into the psychology of decision making has uncovered dozens of cognitive biases that impact our judgment. For example, confirmation bias may lead a therapist to disregard crucial information that would invalidate their original diagnosis and therefore treatment choice.
• Prejudice and stereotypes can be deeply rooted within the human psyche, and try as we might to overcome them, it can be difficult to eliminated them completely. Furthermore, we've all heard of high-profile cases where therapists have used their position of power to take advantage of vulnerable clients. Fortunately these cases are rare, but how often do they go unreported? Humans are susceptible to ethical misconduct in ways that aren't even possible for machines.

Round 2 Summary

Do humans make bad therapists? Of course not. However, I hope that after reading this post and the previous one, you appreciate that the issue is more complex than it may appear. In some areas computerized therapy systems will always fall short of their human counterparts, but in others they have already surpassed them.

________________________________________

Fjola Helgadottir, PhD, is a senior research clinician at the University of Oxford and co-founder of AI-Therapy, a developer of fully automated CBT-based treatment programs. Follow her on twitter @drfjola.