I went to IAMSE this year braced for a dispirited room. AI is accelerating, board-prep pressure keeps getting louder, budgets are tightening, and the climate for higher education and science is just plain discouraging. I expected to find a community running low.
I found the opposite, and it has stayed with me.
This year's theme was "Flourishing Through Change." I left convinced it wasn't just an aspirational theme. It was a fair description of what I saw.
If I had to pick one word for the mood in the rooms this year, it would be ready.
Not reflexive optimism, and not denial. People named their concerns plainly. They talked about struggling to keep up, about feeling underprepared for how fast things are moving. But mostly what came through was a clear-eyed willingness to engage and a mature readiness to move forward.
Kimara Ellefson, who leads strategy and partnerships for the Kern National Network for Flourishing in Health, put it best in her plenary:
Flourishing isn't about how you, internally, are doing. It's about bringing your whole authentic self into community and opening yourself up to the same in others, regardless of where they are coming from, and learning and growing from that interaction. You can only truly flourish in an inclusive and richly diverse environment.
That’s exactly what I experienced, and I think it's the source of the evident readiness. IAMSE generates flourishing because it generates a resilient community. Many will return to institutions where not everyone is as motivated or hopeful. Conferences like this create space to be replenished, and to bring that energy home to help colleagues flourish too.
It seemed the students were everywhere this year, bringing their own ideas and energy into the mix. And it brought something into sharper focus. Faculty and students are at roughly the same place with AI right now. Neither group has it figured out, and that flattened the usual hierarchy into a real spirit of shared exploration.
No surprise, AI was the dominant theme running through everything: teaching and learning approaches, assessment, faculty development, even research methods. The IAMSE community of interest on AI drew so many people the lunch session had to be moved into the ballroom. The vendor banners all promoted AI-supported something.
For me, the volume and hype isn’t the interesting part; rather, it's the discussions that sort signals from noise.
Here's a distinction I keep coming back to. Some of the most visible AI applications are impressive at first blush. What isn't clear is which ones actually promote deeper learning, support educator long-range goals, or will ultimately prove sustainable or effective. Take AI that maps commercially developed content to a school's learning objectives and links it to board-style self-assessment questions. At this stage of AI's maturity, that's a truly useful quick fix. It addresses a real structural problem that otherwise costs faculty and administrators hours to build and maintain, and it feeds students’ test-prep hunger. I understand the appeal. The same goes for tools that promise to help faculty generate their own bespoke content, though in practice the constant prompting and re-prompting adds about as much work as it saves. Both serve the appetite in front of us. Neither builds the conceptual understanding and clinical reasoning clinicians actually need.
What's truly impactful is harder, more careful, and far less flashy. It's AI that augments the real work of teaching, feedback, and assessment, that is created and validated by educator peers, and that grows and sustains through a community model, so faculty can spend real time with students and individualize learning. That's hard to build. It's where Aquifer keeps investing, alongside educators, to strengthen teaching, assessment, and individualized feedback while protecting the human relationships at the center of learning and patient care.
Not every conversation was about AI. A full thread ran through service-learning, health access, and wellbeing: students in free clinics, outreach to migrant communities, food and nutrition initiatives. Those threads remind us what should ground any AI-supported work: the goal is never technology for its own sake, or even for our own sake. Rather, it is grounding all healthcare education in better care for more people.
Competency-based education and programmatic assessment were present, though not clearly dominant the way AI was. Schools know they need to address these. What's missing is the roadmap, especially for the pre-clerkship, basic-science years, where almost no one has a clear answer yet for how competencies apply in a meaningful way, and how this fits into the continuum of student learning.
For me, those two threads — what we teach in the basic sciences, and how we know students are ready — kept colliding around a single question.
What do students actually need to learn about basic sciences in the age of AI?
If a tool can retrieve and synthesize facts instantly, what is the irreducible understanding a clinician has to carry in their own head, and keep building on across a career, now with AI as an ever more capable partner?
I've become convinced the answer is concepts. No longer "What is the dosage of this drug?" and not even "How does this class of drugs work in the body, and when would it fail?" but why? Conceptual understanding is what lets a clinician collaborate with AI rather than be replaced by it or misled by it. It earns them a seat at the table: the standing to question an output, to know when it's wrong, to push back. A bright student isn't ready to walk into a boardroom and make the call simply because they've shown they’re smart. They earn that seat by deeply understanding the business, its context, and the decisions in front of them. It's the same with AI. The clinician has to bring an understanding that demonstrably belongs in the conversation, one that can keep growing as AI's capabilities grow alongside it.
That's also how I've come to think about the arc of what we do at Aquifer: concepts, cases, competencies. Concepts give learners the enduring foundation. Cases give them a place to apply it and build clinical reasoning. Competencies are how we move from "seems prepared" to evidence that they are.
Here's what struck me most: the founding principle of Aquifer has never been more relevant than it felt this year.
Educators remain too stretched to build deeply on their own. Not for lack of talent or will, but because the demands are relentless, budgets are tight, and time is the one thing no one has enough of. I heard this everywhere at IAMSE, and rarely as a complaint. More often it came out wistfully: people wishing they had more time to learn and play with new tools, not less work. That distinction is important, and it points straight at why a consortium model works. When a community of trusted peers does the hard building on everyone's behalf, faculty implement with confidence instead of shouldering it alone.
That's the whole idea, and it's why what we offer can't be replicated from outside this community. Aquifer was created in rooms exactly like the ones at IAMSE, over years of shared commitment to this work. What's new is the chance to seize what AI makes possible, and meet what it demands, to deliver scalable, personalized education even with finite resources.
The future of medical education won't be defined by a single innovation. It's being shaped by educators working together, and with their learners, toward more thoughtful, more individualized, more practical teaching.
The most encouraging conversations I had pointed toward what becomes possible when organizations like IAMSE, Aquifer, and others stop working in parallel and start working together. I came home more optimistic than I've been in a while about what that coordinated effort could unlock.
This community is ready. Ready for AI, ready for competencies, ready for the next chapter.
Flourishing in our field isn't an individual achievement. So let's keep building what comes next, together.