Patients First. Platforms Second.

Why chasing tools before defining problems wastes time, money, and patient trust 

Here’s the point: the patient—not the platform. If a shiny tool doesn’t move a specific decision closer to a better outcome at the bedside, it’s theater. 

Merriam-Webster defines “innovation” as “a new idea, method, or device”.  Wikipedia adds “that offer improvement in offering goods or services”.  So, it’s not just about doing something different but improving our outcomes by doing something different. 

Eroom’s Law is a concept that represents the inverse relationship between investment in technology and the success of drug development.  You saw that right- “inverse” relationship.  Eroom’s Law fairly represents our challenges in translating new ideas, methods, or devices to improved goods or services. 

Take Alzheimer’s as one stark example. Billions have been spent on animal models and AI-driven drug discovery platforms, yet more than 99% of clinical trials in Alzheimer’s disease have failed. That isn’t because the ideas were missing—it’s because we framed the wrong problems, and patients paid the price. 

Innovation in drug development isn’t broken because of a lack of great ideas or tools. It’s broken because too often we mistake noise for signal, hype for strategy, and platforms for solutions. And when that happens, the casualties are not just budgets—they’re patients left waiting.  

The three of us—Brian, Nick, and I—didn’t come together to create yet another soapbox for opinions or predictions. We came together because, from different angles, we’ve all been frustrated by the reality of Eroom’s Law: chasing solutions before we’ve agreed on the problem. 

We also share the same north star: if the work doesn’t change a decision for a patient, it does not count. Brian has lived the hard edges of translational misfires and the ethical imperative to use animals only when it serves patients; Nick has seen where AI clears fog and where it invents it; I’ve carried clinical contradictions back upstream to rebuild models that actually reflect patient reality. 

This newsletter exists to turn those scars into usable guidance—patient-first, problem-first, tool-second. It’s meant to move beyond hype and dogma and build a permanent forum where scars become lessons, where different disciplines meet, and where patient-first thinking is the filter for every story we tell. 

In this issue, we will take you inside the fault lines. Brian Berridge reminds us that without disciplined problem-definition, innovation is just the tail wagging the dog. Nick Kelly punctures the myth of push-button AI medicines, showing where AI truly matters—and where it doesn’t. I take you into the messy middle, where failures, feedback loops, and cultural honesty make or break the future of translational science. Together, these essays challenge the industry: stop chasing silver bullets, start asking harder questions, and anchor every tool to a patient decision. 

That’s the common thread running through this first issue of Innovation to Impact – Ruminations and Ramblings. Brian calls out the danger of letting the “tail wag the dog” when technologies chase ill-defined problems, reminding us why the stakes are human. Nick walks through the promise (and limits) of AI, showing where it can actually help patients—by rescuing stalled programs, tightening safety signals, and bringing real-world quality-of-life evidence into view. I reflect from the “messy middle,” where patient outcomes—not platforms—force us to build feedback loops that reshape our models instead of the other way around. 

We didn’t set out to have these essays harmonize with each other, but the resonance is hard to miss. All three circle around one stubborn truth: progress doesn’t come from the newest gadget. It comes from naming the patient-relevant decision, then choosing the minimum set of methods that can change it with confidence. When we invert that order—when we let tools dictate the questions—we end up with noise, expense, and, worse, false certainty that wastes patient time.

This newsletter is a place to slow down, interrogate assumptions, and tell the stories that never make it into glossy decks. If you’ve stared at trial data that contradicted your “perfect” model, or watched a billion-dollar AI platform fizzle when it faced a real-world patient outcome, you’re among friends. 

So here’s the invitation: don’t just read along. Email us, comment, or submit a note with your own war stories, near-misses, and lessons learned. Tell us where you were wrong and what changed when you put patients first. We’ll highlight selected reflections in future issues. Because innovation isn’t spectacle; it’s the gritty, iterative work of defining the right problem and solving it for someone in a hospital gown, not someone in a slide deck. 

Welcome to Issue One. Let’s keep the ruminations—and the ramblings—pointed where they belong: at decisions that tangibly improve patients’ lives.