Hello X!
If you are a real-world evidence (RWE) / health economics & outcomes researcher (HEOR), I created a reference website. It lists methods, code (R, SAS, Python), examples, visuals, and references for over 300 concepts.
It also has examples and code to generate almost 50 types of visuals.
And some annotated claims and learning tools.
Coming soon AI study analysis tools for protocol, SAP, Table shell, and analytic code outputs. Perhaps also literature abstraction.
All machine readable and accessible.
https://t.co/7WwfSRyQCy
Feedback very welcome! Be kind, but honest, it’s my first public project.
Free to use under MIT license.
@anthropic Claude Fable 5 is killing me with this biology ban. I can’t even do epi methods research with it because god knows calculating the proportion of days covered (med adherence measure) leads straight to bioterrorism. Really?!
I’ve been using @grok code in CLI. The speed is impressive. The presentation is super usable for non-coders. The thinking is also pretty good. I don’t love the internal monologue streaming (I should tell the user xyz; I need to remember qrs)
We are hiring a bunch of Members of the Technical Staff for @GoogleAIStudio who can blend PM, design, eng, and more
If this is you, pls DM me, we will move fast for the best people.
Claude is far better for non coding work. The tooling is just better. Codex is app is the way to build things. If OpenAI can get their harness right and train for knowledge work, it’s over.
@grepdotai I would love to to apply to my scientific use case. I’m trying to recall information from protocols for assessment. Best I’ve gotten with tuned SOTA models is 73%