The first chatbot, ELIZA, was created by MIT’s Joseph Weizenbaum as long ago as 1966.
Named after Eliza Doolittle from George Bernard Shaw’s play Pygmalion, this early conversational software imitated a psychotherapist (Weizenbaum, 1976; Mullins, 2005).
While its intelligence was limited, simply rewording patients’ phrases as questions, Eliza proved incredibly convincing.
Its most famous script, DOCTOR, successfully simulates the renowned psychologist Carl Rogers (ELIZA, n.d.).
Patient: I am unhappy
ELIZA: Can you explain what made you unhappy?
At the time, it was the next step in a much longer journey toward intelligent software and artificial intelligence (AI).
This article explores AI’s recent developments in psychology and its future potential.
Before you read on, we thought you might like to download three Emotional Intelligence Exercises for free. These science-based exercises will not only enhance your ability to understand and work with your emotions, but will also give you the tools to foster the emotional intelligence of your clients, students, or employees.
This Article Contains:
- The Role of Artificial Intelligence in Psychology
- 5 Examples of AI Use in Psychology
- How to Use AI for Psychological Testing
- Using Artificial Intelligence in Cognitive Psychology
- Top 5 AI-Based Psychology Apps
- 5 Artificial Intelligence and Psychology Degrees
- PositivePsychology.com’s Relevant Resources
- A Take-Home Message
The Role of Artificial Intelligence in Psychology
The term AI is typically used to describe both the “technology designed to perform activities that normally require human intelligence” and the multidisciplinary field of science concerned with understanding and developing that technology (Luxton, 2014).
From robots to software distributed across networks, AI is now a commonly used term. And while it can mean different levels of complexity and performance depending on the context and the commentator, it emulates either complex human behavior or specialized intelligent functions.
Most importantly, AI can learn without being explicitly told how to do so.
Psychology, mental health in particular, is one of the more recent areas of focus for AI. As AI lengthens its reach, it is becoming increasingly crucial for psychologists, therapists, and counselors to understand the existing capacity and future potential for the technology to transform mental healthcare.
How can AI help mental health professionals?
Luxton (2014) says that AI can simulate a practitioner, with capabilities beyond its human counterpart.
Indeed, the following examples of advanced technology perceive beyond our human senses to assess patients (Luxton, 2014):
- Infrared imaging to recognize temperature changes
- Facial recognition to confirm a patient’s identity
- Optical sensing to analyze facial expression and eye blinking
- Vocal analysis to perceive subtle differences in speech characteristics
- Olfaction (smell) analysis to identify intoxication
While AI can conduct therapy sessions, e-therapy sessions, and assessments autonomously, it can also assist human practitioners before, during, or after sessions.
Physical assessments such as increased heart rate or temperature changes in response to challenging questions can provide valuable and insightful additional data for the clinician.
Not only that, recording data, managing record keeping, and triggering automatic follow-up actions would free up valuable time for the human professional.
AI as an expert system
Expert systems were one of the first uses for AI within the medical field. While not everyone agrees that expert systems qualify as AI, they undoubtedly assist decision making by combining knowledge and expertise from professionals.
While such systems have been around for several decades, their design has moved from predominantly rule-based logic to making decisions based on data mining and fuzzy logic – the cognitive scientists’ term for handling partial truths (Luxton, 2014).
Through enhancements such as adding speech recognition and natural language processing to expert systems, it is not difficult to imagine technology like Siri, Alexa, or Google Assistant offering therapist-like sessions or expert advice at relatively low costs and without clients having to leave their home.
AI provides other such opportunities. Combining its body of expertise with personal records can monitor health conditions and spot potential contraindications for medical treatments.
Additionally, AI-enabled clinical support systems offer increased capacity, handling a greater volume of highly complex data than their human counterparts could manage and making it available anywhere throughout the day. The burden of time-limited mental health resources can be significantly reduced, and AI can provide more focused engagement with those needing it most and when urgently required (Luxton, 2014).
Computer-generated simulated worlds, known as virtual reality (VR), offer a safe, cost-effective environment for patients to explore their issues. Through immersion, the environment can be made more real for the individual, tailoring circumstances and dialing stressors up or down.
Virtual companions, including digital pets, also promote wellbeing while managing loneliness in easily accessed environments (Luxton, 2014).
Virtual Reality Therapy can be a safe way to deal with post-traumatic stress disorder, and is discussed in our linked article.
Augmented reality overlays the potential flexibility of VR onto the actual world. It uses the readily available processing power of tablets and smartphones to safely connect individuals with the source of their anxiety or personal coaches.
How clients can heal when confronting their fears is also mentioned in our article about Exposure Therapy.
Computer games have successfully increased engagement with reluctant patients and encouraged treatment adherence. By providing a discreet and gamified option for patients, AI-enhanced games can side-step the stigma associated with mental health treatment and provide realistic situations tailored to patients’ needs.
The online computer game Second Life has been successfully trialed as a vehicle for virtual coaching and directed gameplay to enable the patient to practice new skills (Linden Research, 2013; Luxton, 2014).
5 Examples of AI Use in Psychology
AI technology that supplements or even replaces the therapist, counselor, or other mental health professional is not in the realm of science fiction or even the near-future; it is available now.
Detection and Computational Analysis of Psychological Signal
The Detection and Computational Analysis of Psychological Signal project uses machine learning, computer vision, and natural language processing to analyze language, physical gestures, and social signals to identify cues for human distress.
This ground-breaking technology assesses soldiers returning from combat and recognizes those who require further mental health support. In the future, it will combine data captured during face-to-face interviews with information on sleeping, eating, and online behaviors for a complete patient view (Defense Applied Research Projects Agency, 2013).
Computer Science and Artificial Intelligence Laboratory
The Computer Science and Artificial Intelligence Laboratory at Massachusetts Institute of Technology has successfully used AI to analyze digital video and identify subtle changes to an individual’s pulse rate and blood flow, undetectable to the human eye.
While extremely valuable during therapy sessions in uncovering nonverbal cues, it can also monitor trauma patients’ breathing or young babies in distress in hospitals (Hardesty, 2012).
Watson Health, IBM’s AI-enabled analysis tool, is now commercially available and comes loaded with medical literature to serve as both consultant and medical expert.
The incredible aim of this AI is to bring together data, technology, and expertise to stand in for or supplement professional physical and mental healthcare, performing diagnoses and suggesting treatments (IBM, 2020).
The RP-VITA robot has been approved by the U.S. Food and Drug Administration to provide remote communication between healthcare providers and patients. It monitors patients’ wellbeing remotely while accessing their medical records.
The system is multidisciplinary, providing support for psychological, neurological, cardiovascular, and critical care assessments and examinations (InTouch Health, 2020).
Mental Health Diagnostic Expert System
Mental Health Diagnostic Expert System uses advanced AI technology to encode expert knowledge of mental health disorders, which it then uses for diagnoses and proposing treatments.
The AI uses a combination of rule-based and fuzzy logic to understand patients’ needs, agreeing on treatment plans that suit their budgets and are appropriate alongside other health conditions (Masri & Mat Jani, 2012).
Combining the benefits of psychological expertise with AI-enabled technology is having a positive impact on patient treatment and healthcare. With the additional benefits of being cost effective and available remotely, it is likely to develop rapidly.
How to Use AI for Psychological Testing
De Mello and de Souza (2019) explored the potential for AI tools to assist data collection, analysis, testing, and evaluation in mental health.
AI technology offers valuable tools for therapy, combining techniques such as data mining (generating new information from deep analysis of large quantities of data) and expert analysis. AI opens up the potential to diagnose existing and potential problems, test, and confirm predictions and treatments.
When used to understand the data from 707 patients with suicidal tendencies in Greater Santiago, Chile, the AI identified a series of factors associated with suicidal ideation and behavior.
The findings led to a series of preventive interventions for at-risk individuals that reduced the risk of suicide and reinforced “psychological wellbeing, feelings of self-worth, and reasons for living” (Morales et al., 2017).
In 2017, Kravets, Poplavskaya, Lempert, Salnikova, and Medintseva created a model that used fuzzy logic to emulate psychiatric diagnoses. It successfully assessed patients and tested its mental health diagnoses based on incomplete knowledge.
Appropriate AI technology provides the means to piece together fragmented information, build mental models, test their validity, and suggest treatments (de Mello & de Souza, 2019).
Using Artificial Intelligence in Cognitive Psychology
Cognitive psychology attempts to understand cognition’s complexity through research, testing, and building models of how the human mind handles and processes complex information during attention, memory, and perception (Zivony, 2019).
AI and cognitive psychology share similar aims – to understand the nature of intelligent behavior – with the former attempting to build such processes using advanced technology.
And while computational modeling and AI have subtle differences, they are both valuable approaches for understanding the nature of intelligent thinking and providing insights into the growing field of cognitive psychology.
Computational modeling involves “programming computers to model or mimic aspects of human cognitive functioning” (Eysenck & Keane, 2015). Artificial intelligence’s underlying processes, on the other hand, typically bear no resemblance to the mechanisms used by the human brain.
Rather than attempt to develop computational models that help us understand human intelligence, the AI designer’s goal is to produce an outcome that appears intelligent. Such processes do not need to be functionally similar to those of a human.
However, there is one particular model that appears to bridge the gap between the two approaches.
Connectionism was originally inspired by the network of neurons that exists within the brain. “Connectionist models typically consist of interconnected networks of simple units exhibiting learning” and model cognition with no explicit rules (Eysenck & Keane, 2015).
While the brain may be described as a highly complex neural network, and connectionist models have successfully modeled specific human-like processes (such as face recognition), the jury is still out regarding whether such models explain human cognition.
However, deep neural networks, inspired by cognitive psychology theories and methods, have had some success in explaining how children learn labels for objects and offer a great example of the benefits of combining knowledge and expertise from multiple disciplines (Ritter, Barrett, Santoro, & Botvinick, 2017).
Top 5 AI-Based Psychology Apps
While the use of AI in psychology remains a relatively new field, the ubiquity of smartphone technology means that many of us have hardware within easy reach to run the increasing number of AI-inspired psychology apps.
A sample is given below:
Woebot, a 2019 Google Play Award winner, encourages the user to think through situations using tools inspired by Cognitive-Behavioral Therapy (CBT).
The mood tracker then shows the positive changes made over days and weeks.
BioBase uses AI to compile and track stress over several weeks through a series of intelligence function tests.
The app identifies times when stress was highest and offers preemptive deep-breathing exercises.
Find the app in the Google Play Store.
Youper provides a personalized emotional health assistant to help treat stress, anxiety, and depression.
The app uses techniques from several therapies, including CBT and mindfulness, to monitor and improve mental health through a series of brief conversations.
Replika is an AI-powered chatbot that provides an emotional connection and virtual friendship to support people going through depression, anxiety, or troublesome times.
Tess is a web-based mental health chatbot that uses AI to offer the user wellness coping strategies. It promotes resilience through text-based conversations via Facebook messenger, SMS, and web browsers rather than an app.
Website access has to be purchased.
5 Artificial Intelligence and Psychology Degrees
Worldwide, many degrees teach the latest in AI and psychology; however, few integrate the two disciplines.
The following programs are ones we have identified that include elements of both.
- Computational Neuroscience, Cognition and AI MSc, University of Nottingham, United Kingdom. School of Psychology.
This is an interdisciplinary program combining aspects of psychology, mathematics, and computer science.
- Digital Health MSc, Bournemouth University, United Kingdom. Faculty of Science and Technology.
This program teaches students how to personalize and promote digital health through chatbots, wearable technology, and the Internet of Medical Things.
- Artificial Intelligence: Cognitive Science MSc, Vrije Universiteit Amsterdam, The Netherlands. Faculty of Sciences.
This master’s degree uses a multidisciplinary approach to learn about the mind and cognition. Researchers and staff are drawn from various backgrounds including psychology, AI, math, philosophy, and neuroscience.
- Cognitive Science BSc, Carnegie Mellon University, Pennsylvania, United States.
This interdisciplinary degree combines psychology, AI, neuroscience, and philosophy, with the shared goal of understanding intelligence.
- Cognitive Science in Education, BSc, MA, and PhD, Columbia University, New York, United States.
Teaching students to a bachelor’s, master’s, and doctorate level, these programs explore the cognitive mechanisms that underpin learning and thinking. Their goal is to improve educational practices and develop innovative methods built upon new technology.
PositivePsychology.com’s Relevant Resources
We have many tools and worksheets to help users gain greater awareness and understanding of emotions, intelligence, and mindsets. Use them to recognize and further develop your own or your clients’ unique talents.
- The Neuroanatomy of an Emotion explores the functional anatomy of the brain to understand emotional intelligence.
- Reading Facial Expressions of Emotions is a useful exercise for reading others’ emotions through their facial expressions.
- Self-Reflecting on Emotional Intelligence helps to improve stress handling through increased awareness of emotional intelligence.
- The Mindset Survey includes eight items to measure beliefs regarding the malleability of an individual’s abilities and whether a fixed or a growth mindset is being used.
- Monitor your thoughts to discern what type of faulty thinking is triggered. Faulty thinking styles can derail thought processes. Identify your thinking style, and consider whether it is helpful or unhelpful.
A Take-Home Message
AI offers a promising approach to assist and sometimes replace selected practices involved in mental health assessment and treatment (Fiske, Henningsen, & Buyx, 2019).
The technology has the potential to provide new types of treatment (including virtual and augmented reality, and games) and the ability to engage with populations that are difficult to reach or engage with.
Such innovative approaches can also free therapists and mental health professionals’ time and resources to focus on urgent or more specialist care (Fiske et al., 2019).
However, there are inevitable ethical issues. At present, there is limited guidance on the development of such tools or how to integrate them with the work of health professionals, their existing technology and tools, and regulatory frameworks.
Other considerations when implementing AI solutions include understanding and agreeing on the level of human supervision required before, during, and after engaging with clients. At a minimum, assessment or intervention must respect and protect patient confidentiality and autonomy.
Yet, if risks from accidental or intentional misuse and ethical concerns are managed successfully, AI offers a practical approach to treat mental health on a large scale. It also provides the capacity to capture and analyze large amounts of data, with the potential for greater knowledge and understanding of mental health and the efficacy of treatments (Fiske et al., 2019).
We hope you enjoyed reading this article. Don’t forget to download three Emotional Intelligence Exercises for free.
If you wish to learn more about another form of intelligence, our Emotional Intelligence Masterclass© is a six-module emotional intelligence training package for practitioners that contains all the materials you’ll need to become an emotional intelligence expert, helping your clients harness their emotions and cultivate emotional connection in their lives.
- de Mello, F. L., & de Souza, S. A. (2019). Psychotherapy and artificial intelligence: A Proposal for alignment. Frontiers in Psychology, 10.
- Defense Applied Research Projects Agency. (2013). Detection and computational analysis of psychological signals (DCAPS). Retrieved from https://www.darpa.mil/program/detection-and-computational-analysis-of-psychological-signals
- ELIZA. (n.d.). In Wikipedia. Retrieved February 15, 2021, from https://en.wikipedia.org/wiki/ELIZA
- Eysenck, M. W., & Keane, M. T. (2015). Cognitive psychology: A student’s handbook. Psychology Press.
- Fiske, A., Henningsen, P., & Buyx, A. (2019). Your robot therapist will see you now: Ethical implications of embodied artificial intelligence in psychiatry, psychology, and psychotherapy. Journal of Medical Internet Research, 21(5).
- Hardesty, L. (2012, June 22). Researchers amplify variations in video, making the invisible visible. MIT News. Retrieved from https://news.mit.edu/2012/amplifying-invisible-video-0622
- IBM. (2020). Medical coding with IBM Watson. Retrieved from https://www.ibm.com/uk-en/watson-health
- InTouch Health. (2020). RP-VITA robot. Retrieved from https://intouchhealth.com/transforming-team-based-care/?gdprorigin=true
- Kravets, A., Poplavskaya, O., Lempert, L., Salnikova, N., & Medintseva, I. (2017). The development of medical diagnostics module for psychotherapeutic practice. In A. Kravets, M. Shcherbakov, M. Kultsova, & P. Groumpos (Eds.) Creativity in intelligent technologies and data science (pp. 872–883). Springer International Publishing.
- Linden Research. (2013). Second life (Version 1.3.2). Retrieved from http://secondlife.com/
- Luxton, D. D. (2014). Artificial intelligence in psychological practice: Current and future applications and implications. Professional Psychology: Research and Practice, 45(5), 332–339.
- Masri, R. Y., & Mat Jani, H. (2012). Employing artificial intelligence techniques in Mental Health Diagnostic Expert System. 2012 International Conference on Computer & Information Science (ICCIS), 1, 495–499.
- Morales, S., Barros, J., Echávarri, O., García, F., Osses, A., Moya, C., … Tomicic, A. (2017). Acute mental discomfort associated with suicide behavior in a clinical sample of patients with affective disorders: Ascertaining critical variables using artificial intelligence tools. Frontiers in Psychiatry, 8.
- Mullins, J. (2005, April 20). Whatever happened to machines that think? New Scientist. Retrieved February 15, 2021, from https://www.newscientist.com/article/mg18624961-700-whatever-happened-to-machines-that-think/
- Ritter, S., Barrett, D. G. T., Santoro, A., & Botvinick, M. M. (2017). Cognitive psychology for deep neural networks: A shape bias case study. Proceedings of the 34th International Conference on Machine Learning, PMLR 70, 2940–2949.
- Weizenbaum, J. (1976). Computer power and human reason: From judgment to calculation. Freeman.
- Zivony, A. (2019, August 5). What is cognitive psychology? The British Academy. Retrieved February 16, 2021, from https://www.thebritishacademy.ac.uk/blog/what-is-cognitive-psychology/