David Deutsch, whose views are very much aligned with the scientific ideas generally advocated on this forum (except for his use of probabilities)
Specifically, his use of probabilities in thinking about the future, and a massive halo effect around the idea that "progress is good" (including future progress), which makes his thoughts about the future essentially useless, unfortunately. There were other serious errors I pointed out in that review you linked, one on knowability of AI's effect on the world, that makes his thinking on powerful AI suspect as well.
articulate and engaging.
This makes his arguments less likely to be valid than equally persuasive arguments from someone inarticulate with no stage presence.
That is a trivial point, since finding a dull speaker persuasive is very unusual. Would you even spend 30 minutes listening to a boring person?
I have been interested by boring people talking about interesting ideas, yes.
I have also multi-tasked while listening to very boring people talk about moderately interesting ideas.
This can also be downloaded as an iTunes podcast: itpc://www.cbc.ca/podcasting/includes/current.xml (LW doesn't seem to let me link to this).
David Deutsch, whose views are very much aligned with the scientific ideas generally advocated on this forum (except for his use of probabilities), and whose new book The Beginning of Infinity has been recently published, talks about many of the topics of interest to LW. (A flash player might be needed to listen.)
While I find his case for the MWI weak (he mostly says that it explains what we observe better, without going into predictive power or Bayesian reasoning), he is, nonetheless, quite articulate and engaging.