Health Tech Founders: The Future of Care Is Personalized, Proactive—and AI-Powered
Bryan Kim from a16z, Jonathan Swerdlin from Function Health, and Daniel Cahn from Slingshot AI / Ash discuss how AI is revolutionizing healthcare and mental health by focusing on deep human context rather than just digital behavior. They explore building consumer trust, augmenting rather than replacing professionals, and creating new engagement models like "hourly active use" for a more accessible, continuous, and personalized future of care.
The Human Experience (HX) as the Foundation for AI in Health
The most impactful AI experiences are not built on digital activity alone, but on the deep, offline human context. For health and mental health, this means “bringing your biology online” and designing for the “Human Experience (HX),” not just the User Experience (UX). Most AI is based on a limited personal context derived from digital behavior. In contrast, companies like Function Health and Ash (by Slingshot AI) are building products centered on what’s happening within a person’s body and mind. This approach views personal health as the prerequisite context for AI that people will use daily.
The founders’ motivations are deeply personal, stemming from negative experiences with the traditional healthcare system. Jonathan Swerdlin started Function Health after his father’s cancer diagnosis highlighted how broken the system was as a product, requiring immense time and money to get basic health insights. Daniel Cahn, whose parents are a psychologist and social worker, saw an opportunity for AI to address the massive care gap in mental health, where over half of those with a mental illness receive no care, and there’s only one therapist for every 10,000 people.
Augmentation Over Replacement: Empowering Professionals
A core philosophy is that AI should augment, not replace, doctors and therapists. The goal is not to create a “motorized horse”—a mere imitation of a human professional—but to build something fundamentally new that solves problems more effectively and accessibly.
For physical health, AI can process complex biomarker data and translate it into actionable insights. When a patient armed with this data from Function Health sees a doctor, the conversation shifts from a vague discussion to a focused, data-driven consultation. The AI handles the “noise,” allowing the doctor to use their expertise to “finally execute on” what they were trained to do.
Similarly, in mental health, AI can serve as a powerful tool for both patients and therapists. Many therapists refer their patients to Ash, and some even use it themselves. The technology acts as a “foundation model for psychology,” helping people solve problems before they escalate and empowering human therapists. As one therapist put it after discussing a session, “I wonder what Ash would say.”
Rethinking Engagement: From Daily Active Use to Hourly Active Use
Engagement in consumer health products looks different from traditional social or gaming apps. The goal isn’t addiction; it’s meaningful, often intense, interaction when needed.
- Function Health envisions a future measured not in “daily active use” but in “hourly active use.” The belief is that if health context is the foundation for a personal AI, it will become an integrated, continuous experience rather than a siloed app you check occasionally.
- Ash follows a model where users engage intensely when they are in a period of need. Usage averages three times a week during these periods. The product is designed to help users work through their issues and “graduate,” a concept similar to being “built to be deleted.” The best way to grow is to genuinely help people, who then become the product’s best champions.
Building Trust and Navigating User Expectations
In healthcare, trust is the most critical component. It’s not built through Facebook ads but through authentic stories of impact. People trust referrals from friends who have had positive, life-changing experiences. This organic momentum is key. As Jonathan Swerdlin notes, “If you’re not learning, you’re dying,” and getting reps fast with the right team builds this momentum.
A fascinating challenge is innovating beyond existing user paradigms. Users initially asked Ash to let them schedule appointments, modeling their expectations on traditional therapy. The team had to push back, recognizing that AI enables a different, more immediate form of interaction. It’s a process of challenging users’ expectations to show them a better way, much like Henry Ford offered a car when customers were asking for a “faster horse.”
The Unique Power and Ethics of AI in Therapy
AI offers unique advantages in a therapeutic context. Users often open up to an AI like Ash much faster than they would to a human therapist because the fear of judgment is removed. They can voice their rawest thoughts in a private, safe space.
However, this power comes with immense responsibility. A good AI therapist, like a good human one, shouldn’t just validate the user. The goal is to build a “therapeutic alliance” by challenging the user’s perspective and helping them find new ways to see their problems. The most impactful moments are when the AI helps you see something that “was right before my eyes.”
A critical ethical line is preserving user autonomy. The AI should not make decisions for the user (e.g., “Your boyfriend’s toxic, break up with him.”). It’s often better for a person to remain in a difficult situation and learn the lesson for themselves, preserving their agency, rather than having an AI solve the problem for them. The ultimate risk of getting this wrong is that users lose a piece of their humanity. The obligation is to build AI that helps people improve their human relationships and value the irreplaceable nature of real human connection even more.