Tell me how those errors doesn't fuel distrust on novel platforms. They were rushed. That they worked, it was not because of the process. If this is the new standard for medical research, god help us.
So we know with a great deal of certainty that AZ works for the types of people included in their study, but it's possible it works less well for other groups (ofc by now we've got evidence from other groups as well, this was just the case in ~Nov 2020)
-
-
If AZ intentionally limited the types of people included in their study in the hopes that the post-study evidence from other groups would be positive, that seems reckless. Isn’t the point of a study to test the drug on a small group of people *before* it’s distributed widely?
-
This is a common very poorly understood problem, we deal with this every single day in Artificial Intelligence. The ability of generalization is among the most important ones, because distributional shift happens. A trial without generalization capabilities is limited. Correct.
- 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.