A paper can be technically correct, statistically significant, and reproducible and still not generalize. In life science, the hard part is knowing what to trust.
QED Science CTO Meir Maor on why life science needs a validity layer: www.qedscience.com/blog/life-sc...
The biggest dataset in science may be the one that doesn't exist: failed experiments.
Until we capture that knowledge, both scientists and AI systems are learning from a filtered version of reality.
QED started from a simple idea: scientific validity and publication success are not the same thing.
"Fishing expedition."
"Mechanistic link unclear."
"Insufficient preliminary data."
"Scope overly ambitious."
"Limited novelty."
"Lacks meaningful departure from prior work."
At some point you realize: you don't just need funding.
You need a grant friend! 💙
NIH success rates are collapsing. The scariest part isn't the rejections.
Bold ideas aren't disappearing. They arrive watered down. Because swinging big got too costly.
Researchers aren't losing ambition. They're settling.
qed for grants, so your science comes through. qedscience.com
It doesn't have to be this way...
AI agents can already scan the literature, analyze data, and generate hypotheses. Some predict they’ll revolutionize medicine. Others warn they’ll flood science.
Tomorrow, @drmichaellevin.bsky.social
and @odedrechavi.bsky.social discuss this topic.
Register here: us06web.zoom.us/webinar/regi...
Thousands of scientists use QED Science for manuscripts. We recently launched for grants. Meet our new Implementation Agent: it works directly in your document, suggests fixes, and applies changes you approve. Like track changes from a colleague. Join the waitlist now 👉 bit.ly/3Q7a8tS
#GrantReview
We do not understand what we have made.
"You can have a baby, you have no idea how you did it. That shares a lot of symmetries with what's going on now."
@drmichaellevin.bsky.social in conversation with @odedrechavi.bsky.social
Watch the full webinar → www.qedscience.com/blog/how-wil...
The ERC is the Champions League of research. Most don't make it through, not because science is weak, but because the system is overwhelmed. @odedrechavi.bsky.social in @timeshighered.bsky.social on what needs to change.
Because your science should define your career. Not the rejection cycle! 🔁
AI can now read, write, and generate hypotheses.
So what happens to science when machines start participating in it?
@drmichaellevin.bsky.social and @odedrechavi.bsky.social discuss what AI may or may not change.
Live webinar · May 28, 1:30 PM ET
Register: us06web.zoom.us/webinar/regi...
