Making Sense of Fellow Conference Speakers
Dr. Jayaprakash Rajendran MD, Visiting Consultant Psychiatrist in Bristol Priory Well Being Centre
With Clinical Interests: General Adult Psychiatry, Addiction Medicine, Forensic Psychiatry, Dual Diagnosis, Complex Mental Health in Secure Settings and Special Interests in: Literature, Psychiatry and Psychotherapies.
The second day of the annual British Indian Psychiatrists Conference 2025 was held in Birmingham in the first weekend of July 25 and two speakers made a particularly lasting impact through their presentations. Dr. Lena Palaniyappan from Montreal, Canada, delivered “Precision Psychopathology with Language Science: Emerging Opportunities for Improved Detection of Mental Symptoms from Clinical Conversations,” while Professor Femi Oyebode from Birmingham presented “The Importance of Psychopathology in Clinical Psychiatry.” Their contrasting approaches to understanding human suffering in psychiatric practice have left me reflecting on the tension between technological innovation and humanistic tradition in our field.
I’ve been reflecting on Dr. Lena Palaniyappan’s presentation from Montreal, where he spoke with such enthusiasm about making sense of human suffering through artificial intelligence. There’s something both compelling and unsettling about his approach. He began with that fundamental question that haunts all of us in this profession – how do we truly understand another person’s suffering? It’s a question that took him straight to Karl Jaspers and the notion of empathy, of understanding a person in their totality, like observing a flower as a complete gestalt rather than breaking it down into its component parts.
But then Palaniyappan made this interesting turn. He suggested that putting yourself in another person’s shoes is ultimately an idealistic stance, impossible to achieve. And perhaps he’s right about the impossibility of perfect empathy, but this is where my own thinking diverges. When I think about the manic patient producing volumes of speech and movement, seemingly exuberant but fundamentally suffering because their activity lacks inherent meaning and purpose, or the person who’s taken a shot of heroin and lies in the bliss of oblivion, or someone who drinks themselves senseless and lies on the road – these are states I cannot truly inhabit, yet I can understand their suffering from a humanistic perspective. Even when they appear to be in the ultimate pleasures, there’s a suffering that’s understandable because you can’t genuinely put yourself in those shoes of taking something and then lying on the street senselessly. The empathetic understanding comes not from perfect identification but from our shared humanity, our recognition of what it means to be disconnected from meaning and purpose.
Palaniyappan’s solution is elegant in its technical sophistication. He showed us how psychiatric illnesses can be understood as disorders of connectivity across multiple scales – from synapses and circuits at the micro level, through brain networks at the meso scale, to thoughts and behaviours at the manifest scale, and finally interpersonal relations at the social scale. It’s a compelling framework, and his evidence is impressive. Eighty years of research on syntax and schizophrenia, with robust effect sizes across comprehension and production domains, all building toward these sophisticated AI models that can detect disorganized speech with remarkable precision.
The technical architecture he described – ChatGPT as transformer plus training, with its decoder-only language model and masked self-attention, trained on vast corpora of human text – represents something genuinely revolutionary. His five-domain NLP [Natural Language Processing] analysis covering lexical, syntactic, semantic, discourse, and pragmatic elements offers a systematic way to quantify what we’ve always sensed intuitively about disturbed speech patterns. When he showed those colourful spectrophotometric images analysing tone, sentence structure, syntax, and volume, I could see the appeal. Here was suffering made visible, quantifiable, trackable over time.
Yet this is where Dr. Femi Oyebode’s perspective becomes crucial, though Palaniyappan doesn’t directly engage with it. Oyebode’s yearning for doctors to be like expert literary writers in capturing human suffering in ways that honour
the time-tested literary traditions of communicating the incommunicable is profound and necessary. He reminds us that there’s something irreducible about human consciousness, something that may not be captured by even the most sophisticated algorithmic analysis. This literary sensibility, this ability to find language for what seems beyond language, represents the living spirit that carries this ancient art forward. The danger isn’t just that we might lose the art of direct clinical observation, but that we might reduce rich phenomenological experiences to data points, forgetting that the clinician’s presence itself is therapeutic.
The systemic pressures I face daily are real and crushing. The Royal College tells us we shouldn’t take more than two new cases to remain sane, yet reality forces us to see six or seven, leaving our heads all over the place trying to piece them together. In my prison work, managing over 1,300 annual referrals, or maintaining caseloads of 350+ young adults in community services, I’ve felt this cognitive overload intimately. But then I think about my own discomfort with even having a notebook and pen in the room, feeling like these simple tools create distance between myself and the patient. If a teacup in my hand feels like a third presence interfering with the sanctity of the doctor-patient relationship, what does continuous digital monitoring do to that space?
In that context, in my overwhelming clinical reality, Palaniyappan’s technological solutions have appeal – any measures that might help manage these crushing demands seem worth considering. But what if this represents a false economy? If the therapeutic relationship itself is healing, if deep phenomenological understanding prevents misdiagnosis and inappropriate treatment, then efficiency gains from AI might be offset by therapeutic losses. We might be optimizing the wrong variables entirely. Sometimes the most important communications happen in the silences, in the hesitations, in what isn’t said. Sometimes breakthrough moments occur not through analysis but through presence, through the simple act of bearing witness to another’s pain.
The strength of Palaniyappan’s approach lies in its systematic nature and its potential for early detection of subtle changes. His work from Canada shows promise, and their research-driven approach to legitimizing patient consultation recording is methodologically sound. The UK’s more casual use of recording devices in GP consultations, treating them like another Microsoft Office tool, perhaps misses the deeper implications. The spectrophotometric analysis he describes, the ability to visualize the quality, tone, volume, and pauses of speech – this offers genuine insights that could enhance our clinical understanding.
But we must ask whether this enhancement comes at the cost of other ways of knowing, other forms of therapeutic engagement. The Hawthorne effect is real – knowing one is being recorded changes how people communicate. Cultural and linguistic biases in AI systems could perpetuate inequalities in care. Most importantly, we risk falling into what I’d call the reduction fallacy – believing that because we can measure something, we’ve understood it.
I don’t think we need to choose between Palaniyappan’s technological enthusiasm and Oyebode’s literary humanism. Perhaps they represent different layers of understanding. AI could handle the routine documentation and screening, the pattern recognition across large datasets, the tracking of subtle linguistic changes over time. This could free us for deeper phenomenological engagement with complex cases, for the kind of therapeutic presence that can’t be algorithmized, for the literary sensibility that Oyebode champions – the ability to find words for the wordless, to honor the mystery while still providing care.
In my forensic work, in addiction medicine, in the Acute ward and PICU settings where I now work, I’ve seen how the moments that truly matter often occur in the spaces between words. The patient who finally trusts enough to reveal the trauma behind their addiction, the young person who recognizes their own agency in recovery, the breakthrough in understanding that shifts everything – these happen in inefficient moments, in the pauses where humanity meets humanity, in those moments when we become something like the literary witnesses Oyebode envisions.
My concern isn’t that AI will replace human insight – it’s that our enthusiasm for measurable improvement might lead us to optimize away the very inefficiencies where healing happens. The question isn’t whether we can build better diagnostic tools, but whether in doing so we preserve what makes the therapeutic encounter genuinely therapeutic. Perhaps the recording of patient consultations, the analysis of speech patterns, the tracking of linguistic markers over time have their place. But so does the cup of tea shared in silence, the
moment when professional boundaries dissolve into shared humanity, the recognition that sometimes the most profound healing happens not through understanding but through being understood.
Having worked across different psychiatric subspecialties, having managed crushing caseloads that force impossible choices between depth and breadth of care, I recognize the appeal of Palaniyappan’s solutions. But I also know that some of my most impactful interventions have come from moments that would never show up in any algorithmic analysis – the shared laugh that breaks through paranoid isolation, the silence that allows space for shame to be voiced, the simple act of staying present when someone describes the unspeakable. These are the moments that demand the literary sensibility Oyebode speaks of, the ability to honor complexity and contradiction, to hold space for what cannot be reduced to data.
The question Palaniyappan poses remains the right one – how do we make sense of a fellow human being’s suffering? His technological solutions offer one answer, sophisticated and promising. Oyebode’s literary tradition offers another, ancient and enduring. Perhaps wisdom lies not in choosing between them but in holding both, using whatever tools we have while never forgetting that the tool is not the encounter itself. The algorithm might detect patterns we miss, but it takes a human consciousness, perhaps one trained in the literary arts, to transform those patterns into understanding, to find language for the ineffable, to remain present to the mystery of consciousness meeting consciousness.
In the end, perhaps the wisest approach is one of technological humility paired with literary ambition – using AI as a powerful tool while remaining sceptical of its claims to understand human suffering in its fullness, embracing efficiency without abandoning empathy, quantifying what can be quantified while preserving space for what Oyebode reminds us is the eternal human task: finding words for what seems beyond words, bearing witness to suffering with the full resources of language and presence, keeping alive the ancient art of healing through the sacred encounter between one consciousness and another.