Research and standards answer different questions, but both earn trust through disciplined procedure.
AI is starting to help with decisions about people's health, in hospitals, and in pro sports, where teams use data about players' bodies to decide who plays and when. Darwin by Medigram is the system that checks the AI: it scores every decision on six kinds of trustworthiness and keeps a permanent record, so whether the stakes are a patient's care or a player's career, nothing happens unchecked and there's always proof. Designed to work on mobile, where the work happens.
Research asks whether a finding is well supported. Standards ask what requirements diverse institutions and implementers must be able to live under. Consequential AI needs a bridge between the two, plus production evidence showing whether those requirements actually held in operation.
A short reader path into Building Darwin, the full CEO Letter.
Why standards and peer review should not be conflated
TTIC describes standards development, accredited CME and peer review as related process-based trust regimes with different recognized procedures. Scholarship rewards novelty; standards reward durable consensus that competitors, implementers and counsel can live with.
The missing contact surface
The TTIC story notes a substantial AI governance literature that often does not cite documents that actually govern. This creates two communities working the same problem with limited contact.
Keeping the contact surface intact
Scientific evidence and accredited standards answer different questions and are verified through different procedures. One is not a substitute for the other. What often goes missing is the contact surface between them: a path from scientific evidence, through the standards that draw on it, into implementation, and back out again as operational evidence that can inform future requirements. This architecture is built to keep that path intact rather than treating each stage as a one-way handoff.
Why Sherri and the Medigram team
That path holds together today because Sherri Douville has worked directly in standards authorship, architecture, and implementation, and can trace a claim from its scientific or standards basis through to the operational evidence that results. The objective is to make that traceability a repeatable property of the system, not something that requires one person’s memory of how a requirement originated.
What Darwin contributes
Darwin adds versioning, defined behavioral expectations, adversarial verification and durable evidence at the point of decision. That creates an operational record that can be examined alongside published evidence and formal requirements.
Research diligence questions
Which claims are evidence-backed? Which model and instrument versions were used? Are inputs and outputs attributable? Can results be reproduced or re-verified? Are negative findings retained with the positive ones? Does operational evidence feed back into future specifications?
Need the full architecture and evidence?
The CEO Letter contains the detailed technical rationale, verification loop, standards context and diligence path.