So he gave a tech-talk at Google, around 2008, and yes he had lots of graphs and such but during the Q&A session he kept retreating into generalized ideas rather than data. I recall the question about how he came up with his numbers for machines to hold a consciousness as one such exchange.

The impression I certainly got was that his approach is to theorize about something, then design experiments to test out his theory. As opposed to running a bunch of experiments and then figuring out a theory that would explain the collected data.

That said, I've got mad respect for his work and have enjoyed his talks and writings. Your comment though suggests you think 'theorist' is a negative in some connotation? Why is that?

> His tendency to handwave rather than retreat to data? What the heck are you talking about? Have you seen how many graphs he puts in his presentations?

The ability to present something in graph form does not mean that it isn't handwavy bafflegab rather than data.

One need look no further than his absurd joke of a "paradigm shifts"/"countdown to the singularity" graph.