Today we're announcing a partnership with Castify to bring you Less Wrong content in audio form. Castify gets blog content read by professional readers and delivers it to their subscribers as a podcast so that you can listen to Less Wrong on the go. The founders of Castify are big fans of Less Wrong so they're rolling out their beta with some of our content.

 Note: The embedded player (above) isn't live as of this posting, but should be deployed soon.

To see how many people will use this, we're having the entire Mysterious Answers to Mysterious Questions core sequence read and recorded. We thought listening to it would be a great way for new readers to get caught up and for others to check out the quality of Castify's work. We will be adding more Less Wrong content based on community feedback, so let us know which content you'd like to see more of in the comments.

For instance: Which other sequences would you like to listen to? Would there be interest in an ongoing podcast channel for the promoted posts?


New to LessWrong?

New Comment
100 comments, sorted by Click to highlight new comments since: Today at 9:12 AM
Some comments are truncated due to high volume. (⌘F to expand all)Change truncation settings

Another comment prompted by the LW channel page...

you have six months to download the audio in the podcast (once downloaded they’re yours forever)

If one is free to keep the files forever after downloading them once, why not call this "buy" instead of "subscribe" and give people the right to re-download the files at any time? I doubt you'd lose much money from that, and I expect that it might actually net you more sales. A big reason why I'm happy to use various digital stores is the knowledge that I can always re-download the files at any time if I have a hard drive crash, and don't need to worry about one more thing that I should keep backed up. Or I can even go on a spending spree and buy a big bunch of things at once, knowing that I'll only download a couple of them now and can grab the rest whenever I have the time for them. If there was a time limit on when I needed to download my purchases, I'd be a lot less likely to do that, since I might forget about it.

If you call this "subscribe" and give people a limited amount of time to download the stuff, you're making their mental frame of reference to be to other subscription services, which I expe... (read more)

That's pretty convincing. For some reason I had that limited access idea locked in my head. It just stuck for some reason. We will definitely have a chat about changing that very soon.

(Some background: We were going to launch initially with monthly subscriptions, where you get new content every month. In the end, we decided to start with a core sequence and gauge the feedback from everyone. So we had a subscription model "locked in" from early days.)

Update: This limitation and wording has now been removed. However, in our terms of service, we give you a minimum of two years to download the content. We'll keep it up as long as it's technically and financially feasible, or for two years, whichever comes later.

The founders of Castify are big fans of Less Wrong so their rolling out their beta with some of our content.


But seriously, this is great. I'm trying to get into the habit of using podcasts and recorded lectures to make better use of my time, especially while travelling.

I've been lurking on this site for a few months and seeing this in my RSS feed this morning was surprisingly shocking. I guess I just assumed that people trying to be more logical never made this kind of mistake. It was a good reminder that a mistake only invalidates the conclusions drawn from the mistake, so spelling and grammar errors should be pretty low on the list of offenses. It's kind of saddening that this kind of problem draws my attention much quicker than serious logical problems.

To be fair, they're a hell of a lot easier to notice. Although there's probably a signalling issue involved as well - particular kinds of pedantry are good ways of signalling "nerdiness", and I think most LWers try to cultivate that kind of image.
Wow. This is fascinating. You, Ezekiel, are basically saying 'I'm aware that a behavior expressing pedantry like that is a signalling thing, that it specifically signals "nerdiness", and that such a person is trying to 'cultivate an image'." "Oh, and I just did that" ... Presumably you value signaling and cultivating an image with the aim of belonging in a nerdy LessWrong in-group. facepalm What are we becoming? P.S. On an unrelated topic, I think the site founder is wrong about some things. And I just thought you ought to know that I'm such a contrarian :)
In my experience, the people on this site don't perceive signalling as wrong or useless, even when it's superficial. I do not understand why that's so because I perceive most of signalling as a waste of resources and think that cultivating a community which tried to minimize unnecessary signalling would be good.
Which is why everyone should just provide the result of a certified IQ test, just so there's less incentive to signal intellectual superiority, with the lines already drawn. (Heh, that was smart signalling!) (Also, that last sentence.) (And this one?) (Diminishing returns probably.)
Darn it. Even though you are talking explicitly about signaling, I still couldn't help myself from liking it. I also like chaosmis' comment. It expressed what I should have.... Though his comment might also be a sinister meta-signaling-signaling trolling :P God, I hate signaling. (Wait, am I doing it right now?) (Oh shit, and now.) (THERE IS NO RELEASE FROM THE KRAKEN! RUN FOR YOUR LIFE AND NEVER LOOK BACK!!)
You say signaling like it's a bad thing.
6Wei Dai11y
Game theoretic models of signaling show that it can be either socially beneficial or wasteful depending on the details of the situation. It's hard to construct accurate models of signaling in real life, so we can't easily conclude whether any specific instance of it is wasteful. Having said that, I'm curious where you see the most wasteful signaling behavior occurring in this community and if you have any ideas what we can do to minimize them.
People make verbose and lengthy comments instead of short and simple ones. People always speak in a certain type of tone, signalling that they are smart but also that they are Reasonable and they are listening to the points of their opponents. People lace their comments with subtle disclaimers and possible lines of retreat. People take care to use an apologetic tone. I think some of this is a somewhat rational reaction to the amount of nitpicking that happens on this site, which is something that I'm also opposed to. But some of this exists on its own and it shouldn't. I'd prefer it if we just got to the point and stated in the argument as simply as possible. I don't know how to change the norms on this site and don't think any macro-action could do it. Individual people (no, no one specific) just need to relax and to be less personally involved in the site or in the things they say and the arguments that they make. Also, the karma system may or may not be exacerbating this behavior, I'm not sure.

People make verbose and lengthy comments instead of short and simple ones. People always speak in a certain type of tone, signalling that they are smart but also that they are Reasonable and they are listening to the points of their opponents. People lace their comments with subtle disclaimers and possible lines of retreat. People take care to use an apologetic tone.

I'm not sure what the problem with any of these is.

  • Longer comments help reduce the effects of a large inferential distance: on the occasion that my comments tend towards the long, it's because I've noticed that short comments on similar topics tend to not be understood by people. A short comment implies that the person it was directed to could have realized it themselves, they just hadn't put all the pieces together; a long comment also supplies some missing pieces. Given that a good comment is useful not just to the person you're talking with, but everyone else on the site as well, it's generally better to supply more pieces and maximize the extent to which the comment is useful. Of course, you can go overboard with this, but I don't think the comments on LW are excessively long.
  • Everyone speaking in a standardize
... (read more)
1 Length is only good insofar as it adds to meaning. Most length on LessWrong doesn't do that. For example, I can summarize your first point as: I don't think any important information is lost there. I disagree with your assessment of communication practices on LessWrong. 2 I don't think we should react to differences in tone the way that we do. The fact that our community has different norms depending on whether or not you use certain tones is problematic. We should try to minimize the impact that things like tone have. Substantive issues ought to be a priority and they ought to dominate to the point where things like tone barely matter at all. 3 Disclaimers discourage argumentative clash and take extra time to think of beforehand. Simply putting down a disclaimer allows you to marginalize issues that others might have with your post, it makes relevant criticism superficially appear less relevant. A better practice that we should be cultivated is to simply concede things after those things are pointed out. 4 The mindset of lines of retreat seems to stem from the idea that arguments are soldiers meant to defend your social status. Mental lines of retreat might be good but discursive ones are generally a way of avoiding responsibility. 5 Cross apply my above response to your argument about tone. You say that they are good social skills. I agree, given the social norms of this site. But I think those social norms are detrimental to cultivating rationality efficiently and so I want to go about changing the social norms of this site. I don't think so. At best, we've just changed the nature of the game. EDIT: Upon reflection, this last point is basically the essence of my criticism. We've just changed the game to make it more superficially rational, but that is more resource intensive and it masks the underlying mindsets that are bad instead of actually changing them.
8Wei Dai11y
I like to think of possible holes in my arguments before I make them. Sometimes I discover minor holes that don't invalidate the whole argument but do reduce its force. (For example, the argument isn't universally valid but only if XYZ is true, and XYZ seems pretty likely to be true but we can't be sure yet.) Should I not point them out myself? Or are you thinking of some other kinds of disclaimers?
Sometimes these are bad, usually not. It's difficult for me to outline exactly what kind of disclaimers are bad because I think they're bad whenever they do more to prevent the earnest engagement of ideas than to help it, and determining which category specific cases fall in depends a lot on contextual things that I'm having a difficult time describing. I know it when I see it, basically. It's easier for me to ask you to make recourse to your own experiences than it is for me to describe these kind of situations all by myself. Personally, lots of the time when I'm writing comments on LessWrong I spend about 30 seconds thinking up the points I want to go over, and then a couple minutes figuring out how to communicate that message in such a way that it will actually be persuasive to my audience. I feel like I spend much more time here trying to "dress up" my comments in the jargon of the site than I do actually learning things. I expect that many other people feel similarly or at least empathize with and understand my perspective on this.
Length is good insofar as it adds to understanding. Humans are not logically omniscient: if they aren't very close to understanding a concept already, a mere summary of the concept isn't enough to make them understand it. They need examples and supporting evidence. Is this actually a problem on this site? I don't recall seeing disclaimers being much used that way here. It's much easier to change institutions than it is to change people, and it's likewise much faster to get people to adopt social norms than it is to make them change their thinking. "Change mindsets not norms" might work in a closed group, but not on a public site that has new people joining all the time.
It's a big shift, for people to become unaffected by tone. Even if it was possible for community members to make it, it would be exclusive to outsiders, who would be affected by the tone of the discussions and would have trouble participating. Better just to use a tone that encourages good discussions. EDIT: Or to put it another way, it's better to make comments in a tone that causes people to respond more intelligently, then to require them to be inhumanly unresponsive to tone.
I agree but also still think that tone is very overemphasized. We should encourage less reaction to tone instead of taking it as inevitable and a reasonable complaint in response to a comment, which is what I think that we currently do.
Disagree. I support using disclaimers to communicate your point clearly. I don't consider 'argumentative clash' to be an intrinsic good to encourage, especially not argumentative clash about points you could (and should) have conceded already. That distracts from the potential for useful discussion, that actually adds information that hasn't already been considered. Making overgeneralized claims and deliberately refraining from putting in any clarifications or disclaimers is a terrible idea. It encourages arguments that achieve nothing more than force you to retreat to the sane position that you should have presented in the first place. Consider hiding disclaimers away in footnotes so as to not distract from the flow. I oppose your attempt to influence the social norms of this site in this manner. The influence is toxic. That's fucking bullshit! Tone matters. I don't know what the fuck is wrong with you that makes you so oblivious to the actual real world consequences of the proposals you are making but you really need to be slapped down like a bitch for the good of the tribe. Your influence and credibility needs to be crushed so these damn fool ideas of yours get no traction. Seriously, shut the fuck up until you get a clue. YOU ARE DANGEROUS AND IGNORANT. Observers, please grant chaosmosis no social status or leeway when it comes to violating the norms that he (or she, or it) wishes to oppose. Naive fucktard. My claim here is that the tone in the above XML-tagged paragraph would be completely inappropriate and unhelpful in any situation where it was not illustrative. Nevertheless, there is nothing in the actual information conveyed by the insulting and vulgar presentation that isn't entirely reasonable. In terms of "substantive issues" the above paragraph is noble and virtuous. But tone does matter and, as a matter of general policy, that kind of tone should never be accepted, regardless of who the offender is. At a more mild level, where the disrespectful to
Also, you say changing the nature of the game like it's not important. It's like you want to play basketball back before they cut the bottoms out of baskets.
I understand what you're getting at, but what specifically is important about this change? I see the added resource intensity as one thing but that's all I can think of whereas I'm reading your comment as hinting at some more fundamental change that's taking place. (A few seconds later, my thoughts.) One change might be that the goals have shifted. It becomes about status and not about solving problems. Maybe that is what you had in mind? Or something else?
I'd just like to say that your complaints about length are pretty funny in their ironic stupidity.
I said that length was useful insofar as it added to communication. Was I particularly inefficient? I don't think so. As is, it's somewhat ironic, but I think only superficially so because there isn't any real clash between what I claim as ideal and what I engage in (because, again, I think I was efficient). And there's not stupidly there at all, or at least none that I see. You'll need to go into more detail here.
In this community, that depends a lot on the topic of conversation.
Correcting spelling errors doesn't waste many resources. But yeah, the amount of pointless signalling that goes on in the nerd community is kind of worrying. Why do I do it myself? Force of habit, probably. I was the dumbest person in my peer group throughout high school, so I had to consciously cultivate an image that made me worth their attention, which I craved.
I would specify instead: signalling "I care about good communication and avoiding misunderstandings due to poor use of language and syntactical ambiguities/misinterpretations" I've got this idea from I-don't-know-where that this kind of signaling is a useful, cost-efficient sonar ping that'll publicly filter for certain types of people, notably those who care about grammar and those who care about avoiding ambiguities. I think attracting both of those groups is a suitable compromise when the only obvious alternatives are much costlier.

I imagine the less technical the subject matter the more likely it is to be useful to listen to as a podcast. I only listen to podcasts when I'm out on errands and so I don't want to or have to devote much mental energy to get something out of the podcast.

So, by this heuristic, I think the quantum physics sequence is probably out and Yudkowsky's coming of age is probably in.

Yes, we're very cognisant that some content just doesn't work as spoken word. We thought Less Wrong content was a great test case because most articles translate well to spoken word and is pretty unique in that the sequences nicely cluster content for us to offer.
Just to counter your anecdote, I (almost) only listen to podcasts when I'm in an idle waiting situation, or doing routine tasks that require insignificant mental energy. I love to listen to things that make me think, so my mind is as far away from the tedious routines as possible. That said, the content here is "designed" to be read, not listened to. Content that tends to provoke a lot of going back and forth for references and rereading crucial bits multiple times isn't going to work in direct text to audio conversion because of the differences of the media. But forcing the listener to think isn't a problem to me - and I do think (even after trying to compensate for possible cognitive biases) that I represent a significant base of listeners.

Minor interface tweak: most websites have taught me that if I click on the site logo at the top, I'll be taken back to the front page. I was expecting that to work on Castify as well, but it didn't, which delayed me for about half a second before I located the "Home" link. It's no big deal, but many users would probably prefer it if you followed the normal UI convention.

Another good suggestions. That's something I overlook all the time and is pretty easy to fix.
I was just on Castify, and it looks like it has been successfully fixed. Impressive response time.

This lookes like it could be very good. I know some people who have started reading the sequences, were moderately interested, but then stopped as they did not have the time. I am definitely going to recomend this to them. In general, I am a big fan of audiobooks and podcasts for listening to in your spare time, and this combines the greatness of that with the super greatness of the sequences. A big thumbs up from me.

Thanks a lot for the feedback. That's extremely encouraging. We thought the getting the sequences done would be great for example that case and we hope it will thinks more approachable for your busy friends.

Speaking personally, I'm really put off by the payment model. You're presenting this as "$5 for a one-year subscription". Now, if this was "$5 for a one-year subscription to all our Less Wrong content, released regularly on the following schedule", then that would seem fair value for money. On the other hand, if it was "$5 to buy this sequence, and you can buy other sequences once we have them ready", then that would be okay, too. As is, though it's coming across as "$5 to subscribe to this sequence for one-year, plus ... (read more)

Sorry about the confusion over the "subscription" of what is a one-time payment for a Core Sequence. The "subscription" status of this first Core Sequence was a result of the way we originally set up things with PayPal. We are working to change the one-time purchases to a "buy" option not a "subscription". Our goal is to do exactly what you said at the end of your post. We will have a single subscription option where you can subscribe to all new promoted posts from Less Wrong. This will be a monthly recurring payment model. Then there will be a list of Core Sequences available for purchase (like buying an audiobook). You'd buy them individually. We will have some additional core sequences coming out shortly and hope to get the promoted posts subscription option up and running very soon. Thank you for your feedback!
Thanks for the quick response. I figured that you probably wouldn't be trying to do that (it'd be awful for business, for one thing), but from what was written on the site as it stood, I couldn't find any reading of it that said anything else.
Hello again Endovior, just wanted to let you know the changes have been made. Now it's no longer a subscription but a single purchase. Thanks again for the feedback and your patience!
Awesome, was waiting for that to be fixed before sending you monies. Subscribed now.

The free sample is pretty good. The reader is awesome. You've probably got my (monetary) support in the long run, though I'd like to see how it'd work with Kaj's suggestion about the business model. Getting new people to listen to the sequences would be easier if it doesn't look like there's any commitment, and/or if a sequence can be 'gifted' to someone.

The one nitpick or suggestion I might have, however, would be to have slightly longer pauses between main points / topics, to let everything sink in. The pacing of the reader is excellent, but Eliezer's w... (read more)

We definitely hear you. We'll pass that suggestion and positive feedback onto George, he's the reader. The subject material can certainly be dense, you definitely do not want to hit the 2X speed button on your iPhone!

Making the Babyeaters/SuperHappy posts into an audio story might draw new people to the site.

If there was a paywall, that would make it drastically less effective of a recruiting tool. There is a fan-made podcast of HP & the Methods of Rationality that, while less professional, is also free.

What we're really interested in doing moving forward, aside from more sequences, is turning the promoted posts into a podcast. This would be offered as a paid monthly subscription.

We'd love the community's feedback on this.

Edit As suggested by somervta, if anyone is themselves interested in narrating some content, please get in touch with us. We'd love to get people familiar with LW content to help out with the load. When we start adding more channels, we expect our of main challenges to be keeping the quality of the readers high.

Just something that jumped at me from the FAQ:

Do I need to create an account?

Yes. This way we can manage which channels you are subscribed to and allows you to sign in and change your Castify subscription status.

An awkward turn of phrase. Consider replacing "which" with "the" and "allows you to" with "let you". Or rephrase it completely.

It's great to get all this feedback. Answer updated. Thanks.

I realize that you have your own voice actors, but it might not be a bad idea to solicit some help from people already familiar with the whole LessWrong conceptspace. I know we have quite a few members with some excellent recording setups.

That's a good idea. Any suggestions on the best way to go about doing that?
The obvious way would be to put a link in your post, and/or set up a second post asking for volunteers (either way, you should put something in the discussion section also). You could ask anyone who's interested in helping to email you a voice sample of a particular article, specifying the length in the post.

I might be missing something obvious but I'm a bit confused about the subscription length. It says on this page that one has 6 months to download the podcast and that you keep them forever. However, it also says in two places that it is 4.99 for 1 year of access. What significance does the year have?

Your confusion is warranted. I quickly changed the terms to one year based on feedback, but didn't update the description text. Sorry for the confusion.
Hey ahartell, I just wanted to let you know that buying flow has changed. Now the sequences are a single purchase instead of a "subscription". We are still planning on releasing a "subscription" model for all the LessWrong promoted posts in the near future.

After a very short amount of time listening to a Text to Speech voice I now prefer it to almost any narrator. They are very good these days so I won't be making use of this.

Who is your voice of choice? (I prefer Graham for nonfiction material. Paul or Brian for male-protagonist fiction and Lucy for female protagonist fiction.)
Just Neospeech Bridget all around. I like the fact that the voice is female and has a British accent and it seems to be of just all around higher quality than all the available voices except Ivona Amy. Do you find that varying by subject the narrator helps with comprehension? Seems like it would take a bit of extra time.
Seriously? Are you sure you've been comparing good narrators to that TTS voice? For me, a good narrator will win out in an overwhelming majority of cases where I can choose between TTS and a good narrator.
A good narrator is, by definition, superior to a TTS (and as TTS improves, voiceover professionals will have to up their game). But what is superior to a TTS, though, will vary according to the listener. What I want of a good narrator, for example (and I am moved to post this from having heard various storytellers of fiction), is someone who keeps him- or herself out of the matter, and is simply an intermediary, like a newsreader or simultaneous translator. As far as I'm concerned, it's a voice, not a person. I don't want a person chattering in my ear when what I want is the text. The voiceover artist's job is, in fact, to be a better TTS.
Hm, that's interesting, because I generally look for just the opposite in podcasts, particularly fiction. When I read text, the voice in my head emphasizes certain parts and changes tone in response to the content of the text, at a reasonably high level of abstraction (i.e. just looking at the syntax and formatting isn't enough). If a narrator isn't doing that, I have a hard time getting into the reading.
Yeah, seriously. And I have a large amount of experience with different narrators. I find that having one fixed voice aids my comprehension and I don't care that much about how sonorous the voice is. If I could take my pick of a narrators, and some how get a text to speech version of their voice I would pick that and only listen to them to get the effect, but that isn't in the cards.
Interesting. I love how everyone has such different preferences. We're definitely going to stick strictly to the human-narrated content, but there is certainly a growing market of services which can get you your TTS content and we think they have a place too.
Which text to speech program do you use?
Textaloud. I describe my method a bit here. 2 - 5 books a day was an exaggeration, I think. It's usually 2 - 3.

Wow, this is such a great news for me. I listen to a lot of lectures and podcasts.

Castify does not appear to have survived? Are the Sequences still out there someplace in audio format?

0[comment deleted]6y

I see that there's an LW channel subscription page, but there doesn't seem to be any kind of preview that one could use before buying. I presume that there will be one?

Hi there. If you go to the home page and press the green button, you'll hear a sample (you can listen to a whole article). But yes, we are planning to have a sample for each channel. For now, I'll just mention that you can hear audio on the channel page. Thanks. Edit Sample now available on the channel page.
Thanks, I managed to miss that one somehow. It might have been because a green, banner-shaped button was classified by my brain as "graphic / advertisement", so my eyes skipped over it. I don't know if that's a common reaction, but if it is, you might want to make it look different.
That's gonna be bad news for us since all the primary action buttons are green and somewhat banner shaped. LOL.
Hey Kaj, Head to to hear a sample (big green button). Thanks, Rick

This sounds like a great idea. I would be interested to hear "Diseased Thinking about Disease" and the Luminosity sequence in a podcast; also an ongoing podcast for the promoted posts would be pretty cool.