Colin Morelli

VP AI Product at @advocatehealth. Previously @ribbonhealth, @sourcehealth, @ro, @yodle. Healthcare + tech. Opinions are my own.

[Nikhil Krishnan @nikillinit]

To me, Photon is such an obvious thing that should exist in healthcare

Instead of doctors calling a bunch of pharmacies individually and coordinating your prescription, they send the prescription to a wallet the patient controls.

Then the patient can choose which pharmacy to send it to, and can shop around with that prescription.

I usually don't shill portcos like this, but I don't think they get enough shine tbh

[Colin Morelli @ColinMorelli]

@nikillinit Great business. Helps that @ottosipe is also just a good dude.

[Colin Morelli @ColinMorelli]

@GrantHesser ACA plans still have a meaningful adverse selection problem, which will make real innovation in this space harder even if - in theory - the ability to mix catastrophic plans with more dynamic and cash networks is interesting.

[Colin Morelli @ColinMorelli]

@dp_oneill @GrantHesser @joshuakelly Agreed. But there are quite a few smaller regional IPAs that are almost certainly within comfortable M&A range for a well capitalized startup. It'll happen. We're already seeing this in brokerages, and I expect we'll see it in IPAs/MSOs, and in TPAs.

[Colin Morelli @ColinMorelli]

@celinehalioua You are also objectively killing it by any standard, which helps!

[Colin Morelli @ColinMorelli]

@anothercohen Congrats on the weight loss!

[Colin Morelli @ColinMorelli]

@mattzot @StuartBlitz

[Colin Morelli @ColinMorelli]

@chrissyfarr "Premium sponsorships" in the case of Lotus

[Colin Morelli @ColinMorelli]

@StuartBlitz Or buy a house in San Francisco after OpenAI/Anthropic IPO

[Colin Morelli @ColinMorelli]

@nikillinit An absolute banger

[Colin Morelli @ColinMorelli]

@ellenjdasilva Most people don’t know what they want. Research problems not products, then build the thing they didn’t know they needed.

[Colin Morelli @ColinMorelli]

Convincing a large part of the healthcare industry that, actually, physicians should bear risk for total cost of care and not the companies whose sole purpose is to manage population level risk might be the most impressive deflection I’ve seen

[Colin Morelli @ColinMorelli]

@aashnarula @StuartBlitz Delve for Medical Licensing

[Colin Morelli @ColinMorelli]

@seanwdoolan Also amazing. I interpreted a slightly different focus for each but looks like I have a lot of playing to do

[Colin Morelli @ColinMorelli]

This is really, really cool. Incredible work. Huge potential in medicine.

[Faisal Mahmood @AI4Pathology]

📣 We are excited and thrilled to announce APOLLO, a healthcare system-scale multimodal temporal foundation model for virtual patient representations.

Trained on 25 billion clinical events from 7.2 million patients across 33 years and 28 modalities, APOLLO learns a unified atlas of medicine. Turning labs, notes, pathology images, medications, and diagnoses into coherent, computable longitudinal trajectories. APOLLO is disease-agnostic by design, a single model that learns the shared structure underlying human health and disease across every specialty, modality, and stage of care.

The possibilities are enormous: earlier risk prediction, treatment response modeling, clinical trial matching, biomarker discovery, and a new generation of agentic systems built on rich patient representations.

Read the pre-print: https://t.co/IUaWyodcrS
Read our blog post about the work: https://t.co/QOeUHYsrYN

👏 🎉Huge congratulations to Andrew Zhang , @TongDing99, Sophia J. Wagner, and the rest of the team.

[Colin Morelli @ColinMorelli]

@txmedai I’ve never seen such a wide gap between demo capability and real world usage for any technology. It’s still quite challenging to actually safely use AI. Possible, but you really have to invest the time!

[Stuart Blitz @StuartBlitz]

Pretty good new Robinhood feature

[Colin Morelli @ColinMorelli]

Great day to have invested my life savings in struggling footwear brands

[Colin Morelli @ColinMorelli]

This study seems quite poor and I can't even tell what they're trying to say besides attempting to dunk on LLMs?

- Differential evaluation was not probability weighted (which seems far more important than an abstract "did you find every possible explanation")
- Dx accuracy was extremely high (I'd guess higher than average human clinician, but...)
- No human comparison (given that's how people get care today, comparing to that seems like the only reasonable thing to do)

The list goes on, but yikes.