[mdisc]

In an ideal situation for science writing, who someone is, their agenda, etc, shouldn't matter for interpretation right? But here having read the piece without that context, and then having someone point out who the author was really changed my interpretation- for better or for worse.

Many of the points she makes about statistical validity make sense - complex, non linear systems are difficult to predict, models need to be validated on data different from their training sets, etc. So those points, without digging into the real meat of how they're applied to these models specifically -- which she knows lawyers won't, serve to mislead.

I think what's also at the heart of this discussion though is whether what she's writing really is science or if it's opinion/has an agenda. When one puts oneself in the position of writing an overview like this, well then maybe it's one's responsibility to present (as well as complicate) the general consensus rather than present an argument that is not that mainstream as more of a consensus then it really is.