I remember Rob Wiblin speaking about episode length at some point, arguing that longer episodes are really valuable in that they allow a much more in-depth conversation than would otherwise be possible; I agree!, but there is a tradeoff with conciseness and use of time.

The essay by Rob analysing this is Our data suggest people keep listening to podcasts even if they’re very long; the relevant part that responds to your tradeoff remark is (emphasis his)

I’ve looked into our own data and, contrary to the expectations of many, people just keep listening to long episodes, at least so long as they’re good. 

People do indeed drop off as episodes get longer, but two-thirds as many people are still with me between 3h30m and 4h as were with me between 30m and 1h. So the benefit of incrementally longer recordings remains high. ...

Another possible objection: maybe fewer people are willing to start listening to longer episodes? Not as far as we could see (see figure 2). There’s no relationship between episode length and the number of people who start playing it.

That said I do agree with Nick that I wish they tightened up their editing. This seems doable in a way that still gets the benefits Rob mentioned in his essay, like getting to new questions the guest hasn’t been asked before, and the guests easing into the conversation over time as Rob et al build chemistry with them ("I find the best moments on the show are often past the 2h30m mark, when we’re both more likely to be at ease, let our guard down, be authentic and go off script").

I am however wary of using marginal listener acquisition (i.e. listener growth) as the main "steer" for 80K podcast fine-tuning, because of the tyranny of the marginal user, which leads to the enshittification of all once-great products.

I agree, for me they are too long and could be tightened up, there's a lot of less-interesting content that could be edited out even. 1.5 hours for me is around the sweetspot of the best longform podcasts (Tim Ferris, Dwarkesh)

In saying that few EAs who I've complained to about this disagree and love the length, but I'm skeptical how much of that is coupled with it being the original format of "the" ORIGINAL EA podcast.

I wonder how many people who have encountered the podcast in the last couple of years find them too long compared to people who have been into it from the beginning.

You don't need to listen to podcasts as soon as they come out :) 

In fact with most media, you can wait a few weeks/months and then decide whether you actually want to read/watch/listen to it, rather than just defaulting to listening to it because it is new and shiny

In fact since you like Rob Wiblin, you can go and listen to old episodes (from another podcast) that he recommends

Re your suggested solution, the reverse approach might be more implementable? Have a shorter, highlights-version of the episode published on a second channel, possibly with some delay to factor in the extra editing time needed? As main channel episodes have to be more polished than 2nd channel episodes, and I'd personally prefer longer, more in-depth podcasts to be prioritised by being on the main channel.