Therapy 2.0: The science of AI-powered language intervention
By Bill Hudenko and Luis Voloch | Published: 2025-11-04 15:09:00 | Source: MedCity News
Over the past 100 years, our mental model of the typical psychotherapy session has changed little: You might imagine a troubled but relaxed client lying on the couch while a curious therapist takes notes and reflects on how to respond. We expect that with the rapid rise of AI based on large language models (LLM), this image will soon become merely an amusing meme of a less accurate and effective history when helping other humans was more of an art form than a science.
To say that psychotherapy lacks scientific rigor is certainly wrong. Since its inception, clinical scientists have published thousands of studies examining the effectiveness of various psychological treatments and techniques that purport to lead to significant reductions in mental distress. Despite this wealth of scientific knowledge, psychotherapy has also been described as an art because the incredible variability in human behavior requires therapists to rely on frameworks rather than precise rules to manage a myriad of unpredictable situations. Unfortunately, overreliance on the “art form” of psychotherapy has led to a proliferation of ineffective treatments that are not based on scientific evidence, and psychotherapy thus lags scientifically behind other therapeutic areas, such as immunology and oncology, where data-driven decisions prevail. Furthermore, many mental health clinicians rely on intuition or life experience rather than the flexible application of known treatment modalities.
So, what do good therapists do that leads to positive change for those suffering from mental illness? Previously, the answer to this question was largely a mystery. While there is an abundance of evidence to suggest that psychotherapy is better than no treatment at all, and that treatment is often equivalent to the use of medication, the exact words and their optimal combination to effect change have been largely unknown. Fortunately, over the past 10 years, our knowledge of the science of language intervention has improved dramatically due to the rapid growth of text-based psychotherapy, as well as telehealth – especially during the coronavirus. During this time, many large companies began to rely on text care as a new treatment method to provide care to more people, and also made treatment available via telehealth. Although initially doubted as a suitable equivalent to in-person face-to-face interaction, it was soon discovered that both text-based care and teletherapy achieve equivalent therapeutic outcomes in most ways to face-to-face care. This shift has also led to something amazing: a wealth of data on communications between patients and providers – a veritable treasure trove of data for the scientific community to uncover the secrets of how to choose words wisely to produce the right therapeutic outcome for a given patient at a given time. For example, we now know that when people improve, they begin to use language with the future tense rather than with the present or past tense; Knowledge that can now be harnessed to encourage a different attitude that leads to rapid recovery from depression.
With these developments, we are more keenly aware than ever that words matter. The right words are important. The right words matter at the right time.
The subtle “You can do this!” From a parent helping their child acquire a new skill, or a well-timed “I love you” are just two examples of how words can create transformative moments in our lives. This truth is truly incredible – that words can be destructive (“I hate you” or “You’re fired”) as well as healing when used correctly and accurately. In the context of psychotherapy, words are very powerful tools that are harnessed to release potential and relieve mental illness. Their application is incredibly precise, with customization required for each patient (for example, the same message must be tailored to accommodate a different race, age, or living experience), and sophisticated knowledge of a patient’s history is required to understand when a patient is most receptive to hearing the right words. For therapists, using words to assess a problem and reach a conclusion presents a unique challenge. Psychotherapy is actually unique in that it is the only medical field in which spoken language is both a primary diagnostic tool and a primary therapeutic tool.
While psychotherapy is arguably in its true renaissance period due to the above-mentioned developments, we are now beginning a new wave of acceleration in linguistic intervention science: one enabled by the AI-based LLM. Artificial Intelligence and Machine Learning have accelerated the process of learning and detecting linguistic intervention to previously unimaginable levels of accuracy and personalization. We believe this will be more impactful than the progress we have seen so far. Since LLMs are effectively controlled via English (as their programming language), this provides a great opportunity for us to enhance our understanding of how they can help. Through the integrity and careful integration of LLM into the physician environment, we have the ability to reach extreme levels of personalization: the perfect words for the right person at the right time.
In the future, it is quite possible that traditional psychotherapy as we know it will change as a new treatment modality that is routinely coupled with AI-powered technologies. However, therapists are also likely to remain crucial to the optimal recovery of those with mental health issues. Although AI will likely be able to harness the power of choosing the right word to use with the right person at the right time, humans remain ideally suited to collaboratively formulate and implement a “meta” plan with patients that takes into account the overall experience and ultimate desires of the individual – often in ways that are not consciously available to the patient themselves. Thus, a more pressing future situation is likely in which doctors are augmented by artificial intelligence in a way that produces “superdoctors” that we have never encountered before. For example, clinicians will have new insights into diagnosis, increased awareness of risk factors, greater understanding of clients in real time, and more effective and timely interventions that can be prescribed electronically. AI will also likely be used to help clients between sessions make progress and provide 24/7 access to high-quality, meaningful interventions directed by a mental health clinician.
As we all know, no one can predict the future accurately. However, what we believe is almost certain is that psychotherapy has changed and will change forever with the MBA, whether we like it or not. Our hope and belief is that it will be for the better, and it is our duty to build and contribute to this future, rather than avoid it.
Image: Vertigo 3D, Getty Images
Bill Hodenko, Ph.D. He has extensive experience in the fields of mental health and technology. Dr. Hodenko is a licensed psychologist, researcher, and professor who holds a joint faculty appointment at DartmouthDepartment of Psychological and Brain Sciences and Geisel School of Medicine at Dartmouth. His research focuses on using technology to improve mental health service delivery and patient outcomes. He has worked with hundreds of clients and taught thousands of students during his tenure at Dartmouth, Cornell University, and Ithaca College. Dr. Hodenko is also an experienced businessman and is the former CEO of Trust Health Inc. and Voi Inc. and Incente, LLC – all mental health technology startups designed to transform mental health care delivery through technology. Dr. Hodenko is currently the Clinical Director at Jimini Health, a company that uses artificial intelligence to enhance human therapists.
Louis Voloch He is the co-founder and CEO of the company Gemini Health. Prior to joining Jimini Health, Lewis co-founded and served as CTO of Immunai, an AI-driven drug discovery company that achieved a $1 billion valuation with over 140 employees. An MIT alumnus with degrees in mathematics and computer science, Lewis has deep experience in machine learning and biotechnology innovation. Lewis won the best thesis award among all MIT doctoral track students for EECS. His career includes leadership roles at Palantir and ITC, where he led data science and machine learning initiatives. Currently, he also lectures at the Stanford University Graduate School of Business, where he teaches entrepreneurship and management in AI-driven companies. Lewis is passionate about leveraging AI to solve complex challenges in healthcare, from accelerating drug discovery to transforming the delivery of mental health care.
ــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــ
