Mick Lynch

[Mick Lynch @micklynchpdf]

My background is in medical imaging with a lot of effort put into cardiac models. Today I one-shotted this using #Fable5 from @AnthropicAI.

Not anatomically perfect, but amazing how far we've come. See the interactive model at https://t.co/MYB4o9es4w
@_catwu @The_Whole_Daisy

[Mick Lynch @micklynchpdf]

I replaced OpenClaw with Hermes Agent on my personal server and honestly the difference is pretty significant.

Much more stable so far, smoother tool execution, and fewer random failures.

[Mick Lynch @micklynchpdf]

@TheRestHistory 4 days and 9 hours 😅

[Mick Lynch @micklynchpdf]

New video showing the project https://t.co/dmGMY3Hfn2

[Mick Lynch @micklynchpdf]

🏥🤖 The future of healthcare AI - not as single chatbots, but as coordinated teams of specialist agents!

Imagine a hospital where AI agents work together like a medical team, communicating in real-time to deliver better patient care.

This is Health Agents Collective 👇

[Mick Lynch @micklynchpdf]

GitHub Link: https://t.co/nvdVctm7Bj

Get Involved: https://t.co/tt3amsx4nY

[Mick Lynch @micklynchpdf]

#moonshotai

[Mick Lynch @micklynchpdf]

Checking out kimi-k2-thinking in Claude Code

[Mick Lynch @micklynchpdf]

A fun conversation
https://t.co/AWblO4lSak

[Yacine Mahdid @yacinelearning]

if you ever wondered how diffusion models can be used for text generation (like in those blazing fast coding demo) check out @juliarturc latest tutorial

in 24min you'll get the main strategy in D3PM/LLaDA, their inference/training tradeoff and the math intuition behind them

[Mick Lynch @micklynchpdf]

I've been building lots with @claudeai's Claude Code powered by @Zai_org's GLM 4.6 -- one tip I've learned is separating context heavy tasks (like research) into sub-agents.

[Mick Lynch @micklynchpdf]

Coming soon:
- Diagnosis agents
- Insurance verification
- Order generation
- Referral management

All working together as the underlying process management system for entire hospitals. The future of healthcare isn't one AI assistant - it's a team of agents.

[Mick Lynch @micklynchpdf]

Meet the team:
🎯 Orchestration Agent - Master coordinator
🔍 Triage Agent - Patient assessment
📊 FHIR Agent - Clinical data

Using Google's A2A protocol, these agents delegate tasks and collaborate autonomously. A coordinated healthcare AI ecosystem.

[Mick Lynch @micklynchpdf]

I created a simple MCP server to query a FHIR server. IYou can ask it to do things like "give me the list of patients called John Smith over age 55 on the diphenhydramine".
Code: https://t.co/nGjHBPg4Ng

Video: https://t.co/2oF1FPy7Oa

[Mick Lynch @micklynchpdf]

https://t.co/mqTN29zDZ7

[Mick Lynch @micklynchpdf]

ReferWell: Improving Specialist Referrals using AI Agents https://t.co/cEXQtbh0dA

[Mick Lynch @micklynchpdf]

Just received my certificate for Build Multi-Agent Applications - A Bootcamp 🙌

Check out this cohort-based course from @mavenhq 👇

https://t.co/6NNemKvWZs

[Mustafa Suleyman @mustafasuleyman]

This is a must-read. We must not lose sight of the transformational benefits of AI.

[Mick Lynch @micklynchpdf]

@adridder @SAXMKTG @YouTube For me, it's the ability to handle more complex tasks by breaking it down into agents. The manager simply needs to know the expertise if each agent and offloads the task there. As @SAXMKTG said, those specialized agents could have access to specialized tools or knowledge.

[Šimon Podhajský @sim_pod]

.@OpenAI's Swarm is an interesting educational proof-of-concept, but it (predictably) doesn't stand a chance against Microsoft's @pyautogen. That shouldn't stop you from trying it, though! Or at least reading the associated handbook.