What Does 2025 Hold for AI/ML/LLMs (and us) in Digital Engagement?
HNY! Stepping into 2025, AI, machine learning, and LLMs are only beginning to reshape digital engagement, patient support and behaviour change. While these technologies hold immense promise, they also bring unique challenges to reflect on:
-
Nuanced, Hyper-Personalization at Scale: Considering leveraging ML to tailor interventions based on real-time feedback, such as step counts, mood tracking, or health survey responses? Integrating conversational AI to provide nuanced, empathetic, and context-aware support for patients and leveraging AI to identify at-risk individuals early? How about enabling proactive interventions in areas like adherence or mental health?
-
Outbound as a Catalyst: Outbound engagement—via SMS, voice, or notifications—remains pivotal in initiating meaningful interactions. We have seen that a reliance on inbound (for health purposes) still requires timely, context-aware touchpoints, particularly for populations that may not proactively seek support.
-
Trust, Data Bias, Equity: How do we ensure algorithms serve diverse populations equitably, without perpetuating disparities? How might we strike the balance between automated engagement and preserving the human touch in patient interactions.
-
Validation, Safety, Explainability: As machine learning models become more complex, how do we ensure clinicians, researchers, and patients can trust and understand their outputs? How much transparency is “enough” in balancing usability and precision?
We are having these conversations across academic, clinical, and industry partnerships. Requirements around combining outbound engagement strategies with machine learning are driving interventions to ensure a traceable, proactive and adaptive approach, maximizing their impact. These exciting ideas are emerging from open dialogue. How are you thinking about these opportunities and challenges in your own work?
Hoping to stimulate a thought-provoking discussion in the new year!