The ills of psychology and psychotherapy have been the same for years – poorly structured data, a multitude of schools and therapeutic methods, the effectiveness of which is poorly or almost completely unverifiable.
Will the algorithms be able to help the man who created the imperfect, flawed system?
According to a 2016 meta-analysis published in the Psychological Bulletin, which was conducted over the past 50 years, psychology has made no progress in predicting acts of suicide. Subsequent cross-sectional studies laboriously try to sift the grain from the chaff and at the same time show how much pseudoscience is in psychology. As an introduction, it is worth mentioning the famous psychologist who almost has the status of a star – Philipha Zimbardo, whose experiment also did not survive the test of time and turned out to be an ordinary fraud, which was discovered by a French documentary filmmaker.
Despite the crises that psychology is going through, there are also areas where attempts are made to apply algorithms and solutions bordering on the use of artificial intelligence and machine learning. Let’s take a look at the most interesting solutions that scientists have managed to create so far. We will omit historical algorithms, such as ELIZA, which is one of the first intelligent interlocutors-therapists, and focus on the subjective, more advanced three.
Say hi to Ellie, avatar therapist.
Despite the fact that this is an algorithm that is already old, despite the poor graphic interpretation of the therapist, it still makes a big impression. The algorithm analyzes our facial expressions, gestures, voice timbre or eye movement and conducts the conversation as in classical therapy. Studies have shown that veterans (who participated in the pilot) suffering from post-traumatic stress disorder were more likely to answer questions to a digital “human” than to a real therapist. The results achieved by the algorithm were appreciated by clinicians who were impressed by the effectiveness of the algorithm and the openness of patients during the conversation with the avatar.
A much more specialized approach is proposed by Spring Health, which creates algorithms that monitor only mental illnesses.
We are able to predict whether the condition of a given patient will improve after the selected therapy – says Adam Chekroud, co-founder of Spring Health
The Spring Health algorithms are used, for example, by employees of Amazon and Gap Inc. For now, however, this diagnosis consists in selecting appropriate drugs based on questionnaires filled in by the patients themselves.
This process is, of course, largely automated and can boast very high treatment effectiveness, but we are still a little short of artificial intelligence, which would be a full-fledged and effective therapist. For now, the effectiveness of the algorithms depends on the analysis of a sufficiently large set of data. The algorithm analyzes data on several hundred thousand patients, taking into account all their medical data, and on this basis creates a whole network of relationships between the type of patient and the therapy that turned out to be the most effective.
On the basis of such a network of dependencies, algorithms are already able to identify people with mental disorders long before their diagnosis by any human doctor.
So we can risk a statement that in the future machines will be responsible for healing not only our body, but also our mind. The only question is how distant this future is.
– American Psychological Association, After Decades of Research Science Is No Better Able to Predict Suicidal Behaviors.