Healthcare AI's biggest problem isn't capability. It's epistemics. LLMs hallucinate confidently because nobody trained them to say "I don't know."
I'm a peds hospitalist building Maieutic: a clinical reasoning engine that treats UNKNOWN as a first-class answer.
More to come.
You aren't addicted TO scrolling/ vaping/ overworking, you're addicted to anesthesia. This is why your attempts to curb addictions require so much willpower https://t.co/ZhZ2uONbP1
@trq212 I had a great audit session with claude code and we found that we did a lot of circular tooling of skills etc. now we built our memory system to guard against tooling all the time and actually getting stuff done. also off limit times
There’s 100% a kind of mania you can get into with coding agents, where it feels like you’re getting a lot done and working all the time but nothing is really happening.
I think it's often a form of avoidance of something else in life.
the team that aggressively targets this agent-agent environment will be a very successful team. In healthcare we need to ramp up ecosystems.patient agent to provide agent Also we need better agent access to records. And I guess first we need safer agents and more education on how to use them.
☎️ Calling all students:
You can now access hundreds of textbooks directly in Notebook, for no charge. From publishers like S Chand Publishing, OpenStax, and Open Textbook Library— browse, select and upload them as sources to ace all of your upcoming exams (no more excuses, ok?)
@GuleidAmina@GoogleForHealth A competitor to open evidence and chtgpt for physician's. google is also comically behind in agent search. how they let EXA and parallel own them is beyond me.
Introducing the Google AI for Healthcare Professionals Certificate, practical training developed by clinicians at Google to help healthcare professionals use AI responsibly in their daily work.
Dynamic workflows don't have to be hundreds of subagents btw! I use them for any task I want to rerun the same way, even just ~10 subagents
It's a JavaScript file you can edit, so you have full control over the subagents and the order they run in. More predictable than a skill!
very weird thing happening rn
technical people are sad, they think their skills are now useless
non-technical people are desperately wishing they were more technical so they could use agents better
Introducing Gemini 4 Carbon, it's a big jump over Argon which we released just 2 weeks ago.
Gemini 4 Carbon is SOTA on cyber, long horizon tasks, ranks #1 on Artificial Analysis V5, significantly improves on usability in domains like coding, and is more aligned.
I want to coin a new term: Clinician-Signed Specification Iteration (CSSI). A fixed model runs on a versioned spec, and every change is a physician-signed amendment that is gated by eval reruns. Anyone doing this for example for skill development? @AbridgeHQ