"amplifying human bias" seems like a very soft way to put it. One should enquire whether if Palantir itself has any form of temporary access to the data from the companies they consult and whether or not they are allowed to analyse it themselves, even if not store it.
If, yes then we are looking at government enabled Cambridge analytica (in many fields not just politics though) all over again. If they can for instance analyse the data from NHS, insurance companies will be drooling millions over that analysis so that they can maximise their profits by getting a very precise understanding of healthcare need trends.
Yea starts waning after a year but there is a phenomenon called back boost is that if you get reinfected a year later with another variant, this still boosts antibodies to the variant of previous infections/vaccs. I suspect she might not have conditioned to select individuals with no recent infections. Also happens in regular vaccines (maybe less in magnitude than mRNA, that I don't know). Without this memory we would be toast and this lady does not even know this or does not know statistics. I hope she is not in some serious capacity for making healthcare decisions.
I somehow suspect she is not even using data conditioned by infection history and looking at a mix including individuals that might be recently infected with covid.
No they ask for gardening experience. This is a serious gardening volunteering job with designing flower beds and what not. It is not digging dirt and throwing seeds around randomly. Also it is for a dog shelter...
and all the water+energy a small country needs.