EDIT: Yup, “Assumption 1.A sufficiently advanced agent will do at least human-level hypothesis generation”.
This is literally fairy tales for financial people who think they are technical. Read Stanislaw Lem instead; he had better understanding of all this half a century ago.
Another questionable assumption is of course that future tense.
In terms of Clarke's ability to predict the future? I think he proved his prescience enough to take his words seriously.
They're going to need some mobility in order to eliminate humanity.
