Are we holding AI to a higher standard than the humans it's supposed to help?

Dr. AJ Blood, cardiologist, clinical trialist, and co-founder of AIwithCare, joins Steve to discuss what it takes to turn AI-powered patient identification into actual clinical trial enrollment—and why technical accuracy alone isn't enough.

They explore the importance of preserving patient relationships, how to evaluate AI against real-world human performance, and why evidence matters more than promises when introducing new technology into clinical research. AJ also explains why physicians aren't necessarily good at translating clinical problems into software specifications, why healthcare procurement can be brutal for startups, and how remote, team-based care could change chronic disease management.

Plus: a clinical research haunted house, the perils of AI-generated presentations, and AJ's ongoing battle with the furry inhabitants of his historic New England home.

00:00

Meet Dr. AJ Blood

01:16

Finding Eligible Patients vs. Actually Enrolling Them

04:15

AI Outreach, Patient Trust & Clinical Relationships

06:47

Are We Holding AI to a Higher Standard Than Humans?

12:25

The Self-Driving Car Comparison

14:38

What Evidence Should Clinical AI Produce?

15:19

The Clinical Research Haunted House

16:38

How AJ Uses AI at Work

19:10

What Should Physicians Stop Doing?

21:32

Algorithmic Chronic Disease Management

23:33

The Biggest Barrier to Changing Healthcare

25:00

Why Physicians Struggle to Write Software Requirements

27:18

Technology Buyers vs. Actual Users

29:45

The Last Ticketed Event

30:26

AJ's Personal Animal Feud

31:39

Closing

What if the traditional site initiation visit isn't the best way to prepare a site for a complex clinical trial? Dr. Kate Trigg joins Steve to discuss why she believes realistic dry runs could prevent operational problems that conventional training misses—and why sites often don't have access to the technology they need to practice before enrolling their first patient.

Drawing on her experience in academic medical center research operations, Kate also explains the hidden costs that study budgets don't adequately cover, how specialization creates new handoff risks, and why a principal investigator's enthusiasm doesn't always translate into operational feasibility. Plus: the phrase she'd ban from study meetings, using AI to defuse angry emails, and the unexpectedly colorful world of sourdough bread.

00:00

Meet Dr. Kate Trigg

01:50

The Hidden Costs of Running a Study

04:07

Is That Really Just the Cost of Doing Business?

05:12

Why Sites Struggle to Negotiate Budgets

06:48

The $20 Study-Meeting Swear Jar

07:40

When “We've Always Done It This Way” Is Valid

09:10

Sleepy Hollow & Spooky Season

10:20

What Specialization Has Cost Clinical Research

13:04

More Handoffs, More Failure Points

14:38

Can Cross-Training Restore the Benefits of Generalists?

17:00

Simulated Participants & Study Startup Dry Runs

19:46

Should Dry Runs Replace SIVs?

21:31

AI, Source Documents & Angry Emails

24:35

When PI Enthusiasm Meets Operational Reality

28:13

Should Site Operations Have Veto Power?

29:06

Sourdough & the Optimization Icebreaker

31:17

Closing

What happens when clinical trial technology works exactly as requested—but makes the study worse? Jason Spivey, founder of Unifora, joins Steve to unpack the operational mistakes that can turn promising technology into expensive problems, from overwhelming participants with notifications to testing site workflows without anyone who understands how sites actually work.

They also discuss why sponsors, CROs, sites, and vendors struggle to align their incentives, how technology gets purchased without enough scrutiny, and why some integrations consume months of work without solving a real problem. Plus: carbon-copy paper belongs in a museum, Jason's enthusiasm for haunted attractions, and how much he'd pay to escape a particularly unfortunate chair-related curse.

00:00

Meet Jason Spivey & Unifora

04:48

Do Sponsors, Sites & Vendors Really Share the Same Goals?

08:02

Why Collaboration Breaks Down

10:13

The Clinical Research Museum

11:23

Haunted Houses & Halloween Horror

12:51

The $1 Million Chair-Fart Curse

14:10

When Technology Works but the Experience Fails

16:32

Who Owns a Bad Technology Decision?

18:06

The Problem With User Acceptance Testing

20:38

Are We Getting Better at Choosing Technology?

22:53

The Mistakes Sponsors Keep Repeating

26:11

How Jason Uses AI

29:25

When Technology Was the Wrong Purchase

32:51

When to Cut Your Losses

34:47

Closing & Where to Find Jason

What happens when an IRB's responsibility to protect research participants runs into the commercial realities of clinical research? Mike Keens, CEO of Sabai, joins Steve to discuss whether IRBs should flag protocols that are technically acceptable but unreasonable for patients, why certain private equity ownership structures raise difficult questions about independence, and what the industry gets wrong about IRB turnaround times. They also explore how an oversight organization can build a business around telling customers “no,” why Mike uses AI to challenge his own thinking, and whether the best regulatory response might actually be “no, but…”

Plus: Paul Simon, outlawing terrible meeting phrases, and a surprisingly consequential decision about finger guns.

00:00

Meet Mike Keens & Sabai

02:33

Should IRBs Judge Participant Burden?

06:15

AI as a Critical-Thinking Partner

09:28

Paul Simon & Setting Expectations

10:28

One Phrase Mike Would Ban From Research Meetings

12:20

Finger Guns, Bows or Salutes?

13:48

Private Equity & IRB Independence

18:40

When Is a Financial Conflict Too Close?

21:00

What People Get Wrong About IRBs

23:41

Are IRBs Just Rubber Stamps?

25:19

How Do You Build a Business Around Saying No?

28:15

Where to Find Mike & Sabai

Independent research sites are expected to move quickly, adopt new technology, train constantly, recruit successfully—and somehow absorb a lot of the work required to make all of that happen. Alexandria Hull, co-founder and CEO of One of a Kind Clinical Research Center, joins Steve to talk about where small sites have an advantage, what sponsors routinely underestimate in study budgets, and why a site's “yes” doesn't always mean a study is truly feasible. They also get into technology that only works for people who love technology, who's responsible for excessive training, and why sometimes the most productive thing a site can give a sponsor is a very blunt warning.

Plus: AI's dash problem, Harry Potter science, and an alarmingly serious analysis of whether it's preferable to fight a bird every day or be attacked with water by a random child.

00:00 — Meet Alexandria Hull & One of a Kind Clinical Research Center
02:43 — Harry Potter, But Make It Science
03:35 — Where Independent Sites Have an Advantage
05:45 — When Does Staying Independent Stop Making Sense?
07:23 — AI, Hyphens & the Blank-Page Problem
10:31 — How Much Can Better Technology Actually Improve a Site?
14:15 — Fight a Bird or Get Soaked by a Child?
15:56 — What Sponsors Underestimate in Study Budgets
19:09 — When Budget Teams Don't Understand the Work
23:11 — The Side Effects of Running a Clinical Trial
24:22 — When Should a Site Say No to a Study?
27:26 — Are Sites Part of the Technology & Training Problem?

What can a green clinical trial dashboard be hiding? Jin Kim, CEO and founder of Miracle, joins Steve to talk about why apparently stable enrollment can mask operational risk, why clinical research teams keep falling back on spreadsheets despite expensive purpose-built technology, and where AI can actually make clinical operations better. Jin also shares how sitting in on a single ClinOps meeting convinced him to abandon the product he had spent months building and focus instead on the problems study teams were actually trying to solve. Plus: treating SIVs like first dates, the mental lessons Jin carried from competitive archery into startup life, and a surprisingly nuanced position on cucumbers.

00:00 — Meet Jin Kim & Miracle

What Miracle does and the operational problems it's trying to solve.

03:37 — Jin's Professionally Damaging Food Opinion

The complicated relationship between cucumbers and Korean barbecue.

04:49 — When a Green Dashboard Hides Risk

Why steady enrollment numbers don't necessarily mean a healthy study—and why teams should investigate what's working even when nothing appears broken.

09:35 — How Jin Uses AI in Everyday Life

AI, health data, sleep, nutrition and optimizing life as a founder.

11:19 — From the U.S. Archery Team to Startup Founder

What competitive archery taught Jin about pressure, resilience and hearing “no.”

14:56 — Should Sponsors Treat SIVs Like First Dates?

Why sites have options, sponsors compete for their attention, and relationships can change site performance.

17:38 — Why Clinical Trials Keep Falling Back to Excel

Why spreadsheets survive alongside expensive clinical systems—and what they're providing that the official technology stack often isn't.

22:42 — Are We Holding AI to a Higher Standard Than Humans?

Where AI should enter clinical workflows, how to evaluate errors and why the consequences of failure matter.

28:09 — The ClinOps Meeting That Changed Miracle

Jin built a sophisticated product the industry said it needed—and discovered nobody was using it. One look at what study teams were actually struggling with changed the direction of the company.

32:56 — Building Around Real Problems

The lesson Jin carried forward: let go of what you want to build and understand what users are actually trying to accomplish.