This is an awful take, and I hope other readers ignore it, and give the podcast a listen/read.
It's a long podcast, and if you skim the transcript you can see that the discussion doesn't start until much later.
I'm a pretty strong Bayesian, and have heard more than my fair share of vague, hand wavy, and stubborn frequentist arguments against Bayesian statistics. When I see an "Against Bayesianism" rant, I'm already biased against it from seeing so many awful arguments thrown out there, mostly to troll Bayesians.
This is absolutely not one of those. This is a very thoughtful and clearly articulated discussion of the applications of, what Deutsch easily agrees, is a correct statistical methodology to larger epistemological questions.
It is long, so I only had a chance to skim this but it is incredibly obvious that David Deutsch is not "seeking attention", but has very legitimate concerns with the mindless application of Bayesian reasoning to larger epistemological problems in science. I'll certainly be revisiting this later for a closer listen.
So that's the best appraisal you can give of David Deutsch: "Just an old man seeking attention"?
Well, leaving aside the usual embarrassment I feel when it comes to the impromptu nonsense a fair share of HN commentators think it's worthwhile to contribute here when there's a piece of news involving physics or physicists, that's an ageist take without any content whatsoever.
There's more old people who know what they're talking about than young people. That's just the obvious consequence of having been around reading and thinking about stuff more time. You'll notice it eventually because as the song goes, time waits for no one.
Yep, we're getting too impatient and that's not helpful when it comes to think deeply about what we've been taught. Yet that's the most important part of any job IMO. I'll check it out, maybe you're right and he's rambling. That would be surprising to me, which is the reason I replied to your post.
I don't think he's rambling. He isn't sure what familiarity his audience will have with the intellectual underpinnings of his argument, so he recaps those before embarking on the argument proper, in order to make sure the audience can follow.
Granted, he does tell a couple
of anecdotes in the process, but maybe that's his style. I think it's fair to consider impatience implicated here - for what it's worth, when I find myself feeling that way about coverage of stuff I already know but not everyone is guaranteed to, I usually just skip ahead or scroll ahead, checking in here and there, until I hit something on point or that I don't already know. (Usually the second one!)
Impatience is an emotion, and while we can't help much what we feel or how we feel it, we can most of the time treat what we feel as input. Think of it, if you want, like a Datadog alert. How do we handle those? By investigating to understand the root cause and taking whatever action that requires in the context, if any. If we let them drive our behavior directly without taking the time for considered action, we easily risk causing more problems than we're likely to solve.
Granted, I don't entirely love this metaphor, which is no less flawed than any. Maybe too some dork on Twitter will use this as an example of the mechanistic techbro attitude endemic to the diseased discourse of Hacker News comments, or something; it does lend itself somewhat to such misrepresentation.
But despite that lossiness I think it's not wholly without use, because it does point at least vaguely toward a way in which we can manage and make valuable use of even the most unpleasant among our emotions, and one that's served me well over the years since I stumbled upon the concept in some writing or other, I've long since forgotten where.
(I don't think Deutsch was rambling, but I certainly am, in an effort to distract myself from a quite unpleasant facial pain I can't do anything meaningful about until Thursday. Please excuse me.)
I read the transcript and he doesn't seem to be rambling any more than anyone else speaking off the top of their head, rather than writing a carefully edited essay.
Also what mnl said. I hope their comment helps you see your comment in the context of how it would appear to people who read HN with a hostile attitude and look for reasons to reject it (not mnl obviously).
He only started talking about Bayesianism about 29 minutes in to the podcast.
Don't blame him for that... the interviewer asked him all sorts of tangential questions before finally asking him about Bayesianism directly.
But even some of the earlier things he talks about (particularly Popper's objections to inductivism) are actually relevant to his critique of "Bayesian epistemology", which he claims is "a species of inductivism" (which, to his mind, Popper demolished).
That's not the comment I was expecting to see at the top.
I haven't read TFA yet but I popped in to say how much respect I've got for David Deutsch and what an influence he has been on my intellectual development.
