Many COVID-19 cures will go the way of cold fusion & stimulus-triggered acquisition of pluripotency.
Show this thread
When a scientific result of extraordinary potential arises, many scientists will try to reproduce it (and some non-scientists may try to exploit it). If after a while, they still can't, it's a pretty clear sign it was a scam.
-
-
-
If you want to scam your way to a scientific career, prefer releasing mediocre results, as they will avoid scrutiny more easily. Extraordinary results create extraordinary reproduction incentives.
Show this thread
End of conversation
New conversation -
-
-
It’s remarkable how often ML work isn’t reproduced, though. We’ve spent the last several days trying to reproduce biggan. Not only is the original biggan codebase closed source, but all the open source clones seem to have flaws!
-
This Tweet is unavailable.
- Show replies
New conversation -
-
-
But the cash is not returned. :S
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
It’s rarely a scam. With p=.05, false positives happen. Then there could be systematic errors that are hard to spot. The BICEP2/inflation incident is an excellent example. Or the statistical interpretation was flawed. That’s why we have peer reviews.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
You may enjoy John Ioannidis’ papers, such as “Why most published research findings are false,” and “How to make more published research true.”
-
This Tweet is unavailable.
- Show replies
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.