[Previous] Focusing Your Attention Discussion | Home | [Next] New Community Website: Features and Tech

New Community Website Project

This post is for discussing goals and strategy, and high-level planning, for a project (under consideration) to make a new FI community website, with better educational help, organization, forum features, marketing, and monetization.

Alan Forrester would be doing some coding and hopefully other roles. Some other people have expressed some interest too.

Although I'm an experienced Rails developer, I want to focus primarily on where where I have the most expertise and comparative advantage: philosophy (e.g. research and educational content creation), as well as high level strategy.

Related Discussion Topics


Elliot Temple on December 6, 2020

Messages (66)

A candidate name for the new community website is Critical Fallibilism but idk. Should choose a name later.


curi at 4:14 PM on December 6, 2020 | #1 | reply | quote

Things a new site needs:

1. Guided new person learning tour

2. Blog type features

3. Forum features

4. On-site Marketing. Landing pages, SEO, audience targeting, speaking their language and talking their pain points and how we can help

5. Off-site marketing: go to social media and get people (maybe not day 1)

6. Payment features (maybe not day 1)

For payments, I should have ~3 subscriber tiers and allow any custom amount.

Something like 5/30/100 tiers

Things to do based on tier:

- read some private part of site

- post to most parts of site (maybe only monthly open threads for free writing)

- start topics (if we instead allow comments ~everywhere for free. not sure about that)

- learning plan review

- public questions

- private questions

- subscriber only audio, video, articles

- discord roles (version 1: honor system react to assign your role?)

- forum supporter badges

- stream voice chats

- weekly or monthly scheduled stream: watch? ask questions on?

- tutoring (custom reward, not in main tier)

The biggest pricing question is whether I want to sell write access to most of the forum or allow public comments in most places. If I do want to sell it, at what price? Make a tier for it. It’ll have to be the lowest tier. No one will want to be a paid supporter who isn’t allowed to post.

Another big question is how much to sell content via subscription vs. as individual or bundled products. Big stuff like YesNo or CF Course work well as products. What about small stuff like podcasts, a single video, a single article, etc? I think people don’t want to frequently make purchasing decisions. So I shouldn’t sell too many different things. Small things should either go in bundles or be included by subscription.

Having a high subscription tier prices a lot of people out. What is suitable to put in it? Throwing in some small things like podcasts makes sense in medium tier.

Low tier to be a poster (or if not that, to post in certain sections and to start topics). 5-20 is the pricing range i’d consider here. (note: any lower amount is allowed with no features besides like a supporter badge, mb discord role, just to support CF. if ppl wanna support at $1/mo they can.)

Medium tier gets you some non-free content (small, separate things): articles, videos, podcasts, etc. 15-50 is pricing range to consider.

Large tier maybe some more behind the scenes and advanced/non-introductory content here and some semi-personalized attention (e.g. monthly group Q&A/tutoring/lecture scheduled zoom/stream). i think price 100 but i could be wrong and maybe would make more money if it was 50 or 200.

Non-subscription: some big content things like a course where i make multiple things that go together

When people subscribe they will get a big archive. Could limit that in some way or maybe not. Maybe the right policy is that if stuff is timeless great content that should be sold in future, it shouldn’t be a subscriber giveaway? Or I should make a better more polished and complete version. Maybe just say “enjoy the value” and treat it as a big signup perk. I think most people won’t sign up for one month to get archive and then leave. May not be much of a problem there. If it becomes a problem you can do things like you get 1 month of archives per month you’re subscribed – it keeps extending backwards in time.

small/medium/large need different public names. don’t call a subscriber small. standard/big/huge is better… but that’s bad. make the names sound more philosophical or something. want names suitable for being displayed by someone’s name (or on hover or their user profile. or maybe colored name by tier). supporter/gold supporter/donor. learner/TA/master. those are awful. idk don’t need to figure it out now.

Deplatforming risks: payment processor, banking in general, webhost, domain registrar. Maybe cloudflare, AWS. Mailchimp or something to help with sending email. Probably want to integrate newsletter with site in addition to integrating subscribestar and gumroad with site. Oh Discord too but they aren’t that crucial. Partly I want this site to get people to use it more instead of Discord. I think async forums are better. Deplatforming risk is reasonably minimal besides Mailchimp. It’s not like using Twitter, Patreon, Kickstarter, Facebook, Reddit, etc. The site will have politics discussion but I won’t emphasize it and I don’t plan to market re current politics (political philosophy and econ marketing are OK though not a priority – I want to focus more on rationality, learning, thinking methods, etc).

---

a document like this is the kind of thing that could be big subscribers only. behind the scenes. and doesn’t make a great product to sell (would have to just be a bundle/archive of a bunch of unedited stuff).


curi at 4:15 PM on December 6, 2020 | #2 | reply | quote

People aren’t just plain incompetent. They are fragile or *bounded*. I watched political hearings. The people who run it and most of the legislators are competent at asking questions, running the meeting, managing the clock decently well, handling interruptions, following procedures when someone objects or makes a demand, etc. There is skill there. There’s something good about it. You can see them using knowledge instead of just making crap up. And it’s knowledge most people don’t have.

Lots of people are good at something in that sense. They have knowledge. Even scientists who make awful papers are good at something. They just aren’t good at all the things needed for their work. But there are some parts of what they do that they are actually good at – there is some significant knowledge informing how they do that part.

McDonalds workers and truckers have skill and knowledge too. Retail workers and customer service know how to deal with some situations. Not all those workers are skilled. Some are lazy and don’t care. Even they tend to learn a few things eventually. And the better people get pretty good at their job.

People don’t like me because I’ve gotten really good at finding what they’re bad at and highlighting it. They are used to controlled circumstances and highlighting their merits. But I push boundaries, find where their skill stops working, and talk about that.

And I find boundaries for people’s skill at arithmetic, grammar, reading comprehension, logic, etc., which are seen as generic skills all intelligent people are good at. And I show they are bad at those. Because it’s relevant and keeps coming up when I try to have philosophical discussions or any sort of debate. It comes up when they try to learn things and use words (both sending and receiving words).

People also are better if they prepare, take their time, make an artifact (e.g. book, video or blog post) at their own pace, etc. They’re dumber in real time when they can’t selectively put forward the stuff that turned out the best. Real time has advantages though. You find out what people are actually like a bit more with real time interaction. When they only show their highlights it’s misleading about what their life is like. A lot of interaction in general is done via highlights, frequently async. Whenever I read a book, what makes the quality tolerable is that it’s highlights. The author would be dumber in real time.

People want to look good, not be criticized for their boundedness and how their real time is bad. And most people don’t have a very impressive highlight either. Even if they are good at something, it’s usually not very philosophical and they might not even know how to explain it as knowledge they have.

Discussion with fairly quick turnaround (async is fine), sharing highlight ideas (articles you took your time on) and non-discussion learning are all important. People need a mix. They need some critical feedback. They need some more casual chatting. They need some stuff they put effort into. They need some time alone trying to understand things. They shouldn’t focus on just one type of thing. Lots of people like one more than the others and do mostly that. They read a lot but rarely talk. Or they try to debate in chat a lot but won’t read anything.

How do I get/help people to do a better mix with some of each on a weekly basis?

What do I do about people’s inability to read logically/objectively blocking their ability to understand what I’m saying what it isn’t culturally-socially intuitive to them? Just saying things they already understand isn’t going to work. I need to say new things and have them understand that. They need to put work into going step by step and practicing. Worksheets with answer keys could help guide that. *Answer keys could be a good thing to sell.* Worksheets are sellable too, but you can also give away a worksheet then sell the answer key.

---

Is it my fault for talking to ppl being dumb or their fault for posting their dumb at my small forum? Should I act like it’s a big forum when it’s a channel with 3 active ppl? Or should I try to maintain the atmosphere that if you post you’re asking for crit.

It’s hard to act like a situation (big forum) that’s visibly not the case. People interpret lots of actions based on the actual situation they see.

Labelled subforums could help. Serious/casual. Or just repetitively asking people if they want a serious, critical discussion of the thing they just said before engaging. Yeah I should do that. The issue is the ambiguity. If they properly labelled their messages as “careless crap I haven’t thought effectively about and don’t want to” then everyone would know to ignore it. They won’t like being asked to clarify and will talk less. So be it.


curi at 4:18 PM on December 6, 2020 | #3 | reply | quote

Questions to consider for CF goals/design:

- What are focus areas and goals? E.g. is it productive to talk politics? Under what circumstances or in what manner?

- What are people’s pain points that I want to help with?

- What do people need to read/do before discussing, if anything? Depends on type of person?

- Can we categorize target audience into a few main groups? How can landing pages, articles or other things differ by group? (One way is custom landing pages for homeschoolers, relationships, politics, AGI, physics but then funnel them all to the same guided learning tour as everyone else. should more than that be done?)

- what to sell? what to give away? how much to charge?

- what do people get stuck on and why?

- what alienates interested people and stops them from trying to make progress?

- what encourages or inspires people?

- what can be done about people wanting to already be good, having a self-image as one of the smartest people in the room?

- how can people’s dishonesty be engaged with and addressed?

- what do I want from people? money, good ideas, feedback, links worth reading, peers, respect, influence, them to discuss amongst themselves in a way that doesn’t require me to do anything but gives me lots of options for things to comment on that are relevant to my interests?

- how can i provide a quick win to people?

- what sort of concrete results can many people get?

- what BIG concrete results can a few ppl get? e.g. 10% of ppl who try

- how much does making stuff easier for people help vs. you can’t bring down the fountain of youth and they need their own motor, so optimizing on my end isn’t useful, the bottleneck is in them? what ways of helping are actually useful? what do ppl have to do for themselves?

- how can we ask for less of ppl’s attention up front but convert them to higher attention later?

- what search terms do they use?

- what key words/topics interest them? e.g. Objectivism is a potential audience

- what motivates people to take action? to spend money?

- how do you avoid bad customers who are a bad fit? if people invest (time/attention or money) and then get rejected or find out it won’t work, they’re often upset. also bad ppl can post and make the discussions worse

- what can be done about hit and run posts? maybe a separate section for stuff ppl want to follow up on? or two comment types in general, like serious/casual (default to casual)?

- making money breakpoints: i don’t need other income. others don’t need other income so can focus on FI more. can help people with scholarships/grants. can hire editing/art/etc. get rich. impress ppl with the community activity and success.

- community activity breakpoints: enough new ppl frequently enough they can talk with each other, see each other, have other ppl around that they can identify with and connect with more easily. busy enough to seem active and like a medium sized community not just a few high activity posters. seems like lots of ppl are having some success and posting some positive stories and praise. can post a question or discussion prompt and reliably get several answers (that aren’t from ppl who have been around for years) instead of it looking like a dead topic (with enough ppl can do that daily, with fewer can still work weekly).

- should i facecam?

- how to handle making community more friendly while still having criticism and honesty, and pointing out major flaws? one thing is framing competence criticism re mastery and autopilot. i think ppl get offended by the idea they can’t do X when the actual issue is they can’t autopilot it well enough. however lots of times they actually screw it up with conscious attention. though from some pov it’s an error rate of 10% or less. like most of the time they read stuff they think they understand it, so they are pretty good at reading in general. whereas i see they actually can’t read this particular thing even with conscious effort, so they seem grossly incompetent at reading to me. it’s cuz they are used to social communication and they aren’t as good at objective or logical type communication. it’s hard to explain that distinction and relate it to their skillset without being really offensive though.

- what can be done to work with second-handers? liars? evaders? passive people? we need to help at least one of those groups or it’s going to be approximately no one. probably all of those 4 groups. most ppl are all 4 of those… and they aren’t very good at reading, arithmetic, writing, organizing, note taking, managing an activity queue, scheduling, etc., either. and they have a bunch of emotions disrupting stuff and controlling lots of what they do.

- how much should stuff be personalized and 1-on-1 attention type stuff vs. impersonalized digital course and tutorials type stuff?

- stream schedule?


curi at 4:18 PM on December 6, 2020 | #4 | reply | quote

FI has a haters problem b/c ppl think they found a community they can be a big fan of and then it rejects them for being second-handed, leftwing, dishonest, incompetent, etc.

They want to come in an be experts dealing with other experts as peers. And they get rejected as not good enough. Many of them had trouble finding any other community. They rejected a bunch of other communities. And they find one they think is good and impressive, and it rejects them.

FI basically doesn’t hard reject people and say to leave, but people do feel rejected and alienated and quit.

---

What pain points can I help with that people care about?

Philosophy is mostly bad. But you want reason. You want to think well. So you want some better, useful philosophy.

People argue with each other and don’t come to agree. It’s not very effective. They need more organized ways to resolve disagreements.

People struggle to make hard decisions and want more clarity and confidence about it, as well as to make better decisions.

People want to work on AGI in a Popperian way, or just a different way b/c they see the mainstream stuff isn’t working.

Some people want good ideas about the world or whatever topic (econ, psychology, animal rights, physics, etc.). I can provide some of that. But it isn’t the key thing. I want to focus more on how to think and rationality tools. Then they can deal with any topic better themselves. In the context of examples of applying rationality stuff – showing how to use it – it’s good to talk other stuff, e.g. to show how to actually resolve a debate about a controversial issue. But I don’t want to be handing people answers. I want them to think for themselves. Even with philosophy itself it’s important to put a lot of work into enabling people to understand things for themselves instead of just telling them conclusions.


curi at 4:18 PM on December 6, 2020 | #5 | reply | quote

Some reasons for new stuff:

- FI community not growing

- revamp things

- reconsider some old community organization/design things

- change (improve) some discussion norms and policies

- partial blank slate

- i kinda inherited the community plus i know more now. lots of room for improvement

- make more money (and be better able to hire ppl to edit videos and other nice stuff). capitalism and commerce are good things! my coding work isn't a big distraction but does take some time. some other ppl are more limited by their jobs and if this grows it could support multiple ppl

- grow community size and help more ppl with rationality/philosophy/etc

- better tech features, esp for facilitating discussion. basically modern ppl won't use email and ppl can't start their own topics here (and this is a Rails 2 website with sketchy markdown related code that isn't in a good position to implement all the nice things a forum could have)


curi at 4:24 PM on December 6, 2020 | #6 | reply | quote

Goals (personal):

- my work, my way – make the philosophy stuff i wanna make

- develop better philosophy ideas

- share ideas so others can learn and give feedback

- be open to debate and criticism

i'd also like to fix things so people don't have bad experiences but without sacrificing criticism, integrity, rejection of some social norms, etc.

growing the community would be good for providing more options for me to skim and pick and choose what to engage with. there's limited relevant philosophy material anywhere else to engage with. also can have more ppl to sell stuff to. and improving the world would be nice. if ppl would learn to think effectively that'd make lots of things better. i'm not counting on that happening though. similarly, it'd be nice to find/meet some smart ppl or peers or something, but i don't expect it.


curi at 4:28 PM on December 6, 2020 | #7 | reply | quote

Risks:

It's hard for me to do everything (philosophy + coding + marketing + business). I want to focus primarily on the philosophy role.

So the main risk is other people letting me down. They could quit FI and become hostile, or quit in a passive/disinterested way, or not do much re the community website project.

So far in the past I have had poor results getting other people to do much to help.

There are also risks re maybe it'll never be popular and never make significant money. OTOH i think it has the potential to at least be medium popular and make substantial money. I think there's a ton of world-class and unique value in my ideas.

Atlas Shrugged:

> “John Galt spent years looking for it. He crossed oceans, and he crossed deserts, and he went down into forgotten mines, miles under the earth. But he found it on the top of a mountain. It took him ten years to climb that mountain. It broke every bone in his body, it tore the skin off his hands, it made him lose his home, his name, his love. But he climbed it. He found the fountain of youth, which he wanted to bring down to men. Only he never came back.”

> “Why didn’t he?” she asked.

> “Because he found that it couldn’t be brought down.”

This is a major risk. I can't think for other people. They have to think for themselves to understand and get value from great ideas. Just making the ideas great isn't enough. I can't choose whether other people think or think effectively enough to learn and use ideas.

Another risk is attracting haters and being cancelled/deplatformed or harassed. I've already experienced quite a bit of targeted harassment and hatred (much, much more than average for FI's community size). Getting significant public attention while having unpopular ideas has some dangers.


curi at 4:34 PM on December 6, 2020 | #8 | reply | quote

#2

> small/medium/large need different public names. don’t call a subscriber small. standard/big/huge is better… but that’s bad. make the names sound more philosophical or something. want names suitable for being displayed by someone’s name (or on hover or their user profile. or maybe colored name by tier). supporter/gold supporter/donor. learner/TA/master. those are awful. idk don’t need to figure it out now.

My suggestion is badges. Some badges you get by paying. Some badges you get by doing things. A person could have 0 badges, 1 badge, 10 badges, etc. depending on what they pay and what they do.

Someone who subscribes could get a colored subscriber badge: bronze, silver, gold.

Someone who buys X products (or $X worth of products) could get a scholar badge.

Someone who has published a learning plan (LP) and monthly progress this month could get the "learner" badge - linked to their LP. I don't think that's awful on its own. I agree TA and master are awful (don't recommend those badges).

Someone who has published a paths forward policy could get a "PF" badge - linked to their PF.

Someone who has posted answers to 25 newbie questions could get an ambassador badge.

etc.

I don't like the designation of "supporter" or "donor" since it implies charity rather than value trading.


Andy Dufresne at 6:34 AM on December 7, 2020 | #9 | reply | quote

#2

> Deplatforming risks

The other risk is lawsuits. Forming an entity to own the IP that's separate from the entity that runs the web site allows those things to be defended separately and also protects the people involved from personal harassment.


Andy Dufresne at 6:53 AM on December 7, 2020 | #10 | reply | quote

#10 I should have said "Another risk is lawsuits". There are probably other deplatforming risks.


Andy Dufresne at 6:54 AM on December 7, 2020 | #11 | reply | quote

#3

> Is it my fault for talking to ppl being dumb or their fault for posting their dumb at my small forum? Should I act like it’s a big forum when it’s a channel with 3 active ppl? Or should I try to maintain the atmosphere that if you post you’re asking for crit.

Posting dumb stuff is clearly the fault of the poster.

The issue for your forum is how to effectively deal with people posting dumb stuff. I don't have a good answer other than I think different approaches are best for different people and dumb posts.

- Some dumb posts should be ignored

- Some dumb posts should be corrected and discussed with the poster

- Some dumb posts should be corrected and nothing further said about them

- Some dumb posts should get the poster banned

- Some dumb posts should be corrected and discussed within the community but not with the poster

- Some dumb posts should be linked to canonical documents (intro, rules, etc.) for correction rather than addressed with new comments


Andy Dufresne at 7:15 AM on December 7, 2020 | #12 | reply | quote

serious vs. casual

#3

Emphasis added:

> Is it my fault for talking to ppl being dumb or their fault for posting their dumb at my small forum? Should I act like it’s a big forum when it’s a channel with 3 active ppl? *Or should I try to maintain the atmosphere that if you post you’re asking for crit.*

>

> It’s hard to act like a situation (big forum) that’s visibly not the case. People interpret lots of actions based on the actual situation they see.

>

> *Labelled subforums could help. Serious/casual.* Or just repetitively asking people if they want a serious, critical discussion of the thing they just said before engaging. Yeah I should do that. The issue is the ambiguity. If they properly labelled their messages as “careless crap I haven’t thought effectively about and don’t want to” then everyone would know to ignore it. They won’t like being asked to clarify and will talk less. So be it.

I like the idea of labeled subforums that are more casual or more serious.

I imagine I'd like one place with little to no criticism, most places with criticism being welcome, and one place to post things you especially want criticism of. But maybe different ratios would be better.

You could encourage new people to only post in the no-criticism area to start with and then to gradually post more in the criticism-is-welcome areas. Or maybe that would be bad because it would reinforce the idea that criticism is scary.


Anne B at 9:31 AM on December 7, 2020 | #13 | reply | quote

#3

> People don’t like me because I’ve gotten really good at finding what they’re bad at and highlighting it. They are used to controlled circumstances and highlighting their merits. But I push boundaries, find where their skill stops working, and talk about that.

> And I find boundaries for people’s skill at arithmetic, grammar, reading comprehension, logic, etc., which are seen as generic skills all intelligent people are good at. And I show they are bad at those. Because it’s relevant and keeps coming up when I try to have philosophical discussions or any sort of debate. It comes up when they try to learn things and use words (both sending and receiving words).

I have a history here of interpreting valid and truthful criticism of my skill at something as meaning “you’re bad at this thing *and* you’re bad at everything and there’s no hope for you”. I’d benefit from criticism more and be happier if I could see it as “you’re not as good at this thing as you thought you were. lots of other people aren’t good at it either. you are good at other things. you can get better at this thing now that you see where the problem is.” I think this is the case for other people too and often a major factor in people leaving FI. But I don’t know how to help new people not hate FI or you over it. It would be silly to give a disclaimer with every criticism, and different people might need different kinds of disclaimers anyway.


Anne B at 10:08 AM on December 7, 2020 | #14 | reply | quote

#2

> small/medium/large need different public names. don’t call a subscriber small. standard/big/huge is better… but that’s bad. make the names sound more philosophical or something. want names suitable for being displayed by someone’s name (or on hover or their user profile. or maybe colored name by tier). supporter/gold supporter/donor. learner/TA/master. those are awful. idk don’t need to figure it out now.

“learner/TA/master” is bad because it equates how much money someone gives to FI with how accomplished they are at philosophy. It also implies that learner and master are mutually exclusive.


Anne B at 10:23 AM on December 7, 2020 | #15 | reply | quote

It’s hard to tell people “I am criticizing your autopilot”. It feels like an insult. They aren’t used to thinking of themselves as acting mostly by subconscious software, not conscious reasons. Often they feel like they are consciously thinking, not in NPC mode. But even when consciously thinking most of the knowledge being used is from autopilot.

It’s hard to help people who only post on their best behavior or in high effort mode. Because then they go back to their life and are in regular mode, use lots of autopilots, and think and act differently. If they only do high effort posts then they’re hiding what most of their life is like from criticism, which generally means it stays pretty bad and the philosophy advice doesn’t change much for them.

Are people competent or incompetent at reading? It’s more complicated. They have flawed autopilots that lead to lots of mistakes in some situations like philosophy discussions. The autopilots are optimized for social communication not objective/logical communication. This is a hard thing for people to consider or face. And they think they can just override that and use conscious analysis in discussions, but they can’t. Their high effort, high attention mode still relies on lots of autopilots of small reading tasks which introduce lots of errors.

---

Wanting criticism should be an opt-in tag. Maybe two versions: some criticism and full criticism. People don’t understand that full criticism includes things they regard as off topic or tangential. Full criticism means unbounded criticism. It doesn’t care about what you think the boundaries of the field, topic, project or interest are. It cares only what is the goal(s) and what is logically relevant to the goal(s). That often means that your scheduling and time allocation policies are relevant, as well as your reading precision and your background knowledge.

I think everyone but me shouldn’t have an unbounded criticism tag enabled by default.

It could be three tags: unbounded criticism, criticism bounded by topic, and criticism bounded by amount. (You can pick both the topic and amount bounds to mean it should be bounded both ways.)

However, maybe it’s better to assume people want *some* criticism or they wouldn’t be at the forum at all, so just the unbounded criticism tag can differentiate. There could also be a “very little criticism” tag.

There could be tags like “small discussion” or “narrow discussion” (low topic branching).

It’s really important to limit clutter and not have too many options. Which options are really worth having? It’s fine to brainstorm lots of stuff but the actual design needs focus. Having a fair amount (e.g. 20) pre-set tags people can use is possible but it takes up a lot of the budget for forum stuff.

Related, re tags and subforums, there could be some main tags which are essentially the sub-forums, and every post must have at least one. If you don’t want to pick one, that’s fine, you get the Other tag as a default.

There don’t necessarily need to be subforums or main category tags. If they exist there are different ways to organize them. I don’t want to do them by topic in general. Physics, programming, economics, political philosophy, politics, psychology, psychiatry, parenting, education, relationships, business, history … no. Having a “religion and politics” topic or a “technical details” area to segregate those could be OK.

could have sections like abstract theory and applications. or learning vs. cutting edge. or talking about books you read. eh that sound more like optional tags at most.

what is the purpose of sections? to help readers find stuff. find stuff when/how/why? Some main cases:

- trying to look for info about a particular topic. main solution: search features

- they like a particular author. solutions: author pages and search

- they read the forum regularly and want to read new stuff. solution: some way to find new stuff

- they are visiting the site for the first time or infrequently. solution: they can either look around at whatever new stuff or use search

topic subforums i think are mostly for people who want some topics and not others *and* regularly read the forum. cuz if you’re picky about topics and not a regular reader, use search.

another use of subforums is when you have high activity, so e.g. you have several pages of threads sorted by recent activity in the last hour.

here are Less Wrong’s major categories:

> Rationality, AI, World Modeling, World Optimization, Practical, Community, Personal Blog

overall i don’t like that. i see some value in splitting off AI b/c some of that stuff isn’t of general interest.

they are sorta doing a theory/applied split but with 3 types of applied. rationality is theory. world modeling means understanding the world. world optimization means improving the world. and practical means dealing with the world. community might sound like a 4th applied but i think that means meta discussion about the community itself (both its policies and place in the world). so e.g. moderation policies and meetup groups go there. and personal blog i think is essentially and Other category where u can write whatever you want.

AI is IMO better done as a “technical” area like i made on discord. So could have: main (philosophy theory and applications together – don’t split those), technical (not general interest), meta, other. This looks reasonably useful to me. Then most people can read main (maybe name it philosophy or rationality but the problem with that is people won’t expect that econ or parenting goes there) and the other 3 sections help exclude stuff most people don’t want to read. Technical excludes stuff like math details, meta excludes stuff like forum policies, community building plans, and stuff ppl wouldn’t see as discussion of topical ideas, and then the other category lets people write whatever they want that they think is a bit off topic, e.g. discussing cooking or gaming (which they can also put in main if they are trying to approach it in a thoughtful way and connect it to any philosophy ideas or methods). More specific stuff like politics or coronavirus could be a more minor tag that people could filter out. Also another major subforum (in the sense of easy to find and browse) could be my posts. Another one could be curated posts: promote the best posts with a special tag.

There are other ways things could be organized, e.g. a thread could be an article with comments, or it could be for a debate or discussion between specific people, or it could be looking for debate or discussion with anyone. Maybe we could auto-categorize by length of the first post in a thread + allow manual override. People might want to filter for long stuff that’s more article-like instead of like “ok gonna debate with joe about minimum wage below”. Actually length-based search could be useful for comments too. Maybe I’d like to read all the new comments that are at least 500 words and that’d be a good way to find the most interesting ones. You could do a search like “comments since last time i visited, and by anyone over 500 words or by curi over 50 words”.

Anyway this setup is sounding OK to me: Main/Details/Meta/Other aka Philosophy/Technical/Community/Off Topic. Feedback? So front page of forum you’d see those four sections plus a few other things like a search option and some suggested/standard searches like my posts, recent long posts, whatever we find is most useful.

There are two different sorts of search. There’s the search where you type in some words and the search where you don’t and you’re just *filtering*, e.g. filter by author, date, length and being the first post in a thread, and then get *all* posts meeting those criteria sorted in some order. That’s different than typing in some search terms. Stuff like an author page for anyone on the site, which can serve as a sort of personal blog, is a filter. Idk if the search vs. filter terminology is good but filters are what i’ve had in mind. Keyword search should exist and is good for finding threads on specific topics, e.g. discussion about Greece or intelligence. viewing posts with particular tags is filtering. keyword search should allow filters but filtering without writing any keywords should be a common thing. subforums are essentially just filters that are being given some extra emphasis as defaults. you could easily have a way to stop filtering by subforum somewhere but have them exist as the default.

it’s hard to get people to do much tagging. like if there are optional topic tags like politics and physics, people will often neglect to use them. making people select one of the four main tags to start a new thread i think is reasonable, not too much work to ask of people. you shouldn’t have to pick any tags to write a comment but starting a thread can be a bit higher effort. and picking one of those four things should be pretty easy. getting people to segregate politics – cuz it triggers lots of ppl and can lead to flamewars – is harder and maybe not worth trying much. if we get big one day we could hire a moderator to add topic tags on all new threads so then politics could actually be possible to filter out (for ppl who want to stay away from it) without users having to worry about doing extra work. btw i wish Daring Fireball had either a way to filter politics out or a way to filter to only Apple-related. cuz i only like his Apple related posts, and i actually dislike some of the politics (as against posts about whiskey, baseball and movies which are neither interesting nor offensive).

tags (or just filterable attributes) that can be done automatically are nice. like author, date, reply count, images?, subheadings?, activity level (some algorithm about how much a thread has been posted to recently), nesting amount (both how nested is this and how much nesting exists below it) and length (if we split length into a few standard categories instead of expecting ppl to type in how much words or characters they want posts longer than – or can have both, a few standard categories but a custom search feature somewhere).

Internet meme pics should go in Other. That’s another reason for having a few main subforums. Don’t wanna ban those but don’t want them mixed with the main serious writing stuff.

Different way to view the four subforum ideas is: general interest (for people in this philosophical community), and then 3 types of not general interest, which are 1) requires some kinda specific expertise 2) stuff about the site/community itself 3) all other stuff that isn’t general interest

i’m not very picky. if ppl wanna talk physics in main that’s ok, but if it’s really mathy/technical then it probably shouldn’t be there cuz most ppl will have a bad time. talking about how to have a good discussion should go in main. but why Joe is banned for a week should not, nor should discussing our plan to recruit Objectivists.

Main: Most users on the site might be interested in this

Technical: Only a minority have the background to engage

Meta: Mostly only the people directly involved or who care a lot about the community would be interested

Other: Stuff the majority might not care about


curi at 12:50 PM on December 7, 2020 | #16 | reply | quote

Pay to Read? Pay to Post? Pricing?

**Big question:** should the forum be around 95% pay-to-post or 25%? And what price to post (probably in $5-20/mo range)?

I see it as two main options: either not much allows free posting (maybe just one public open thread created per month) or a lot has free posting but there’d be a subscribers-only section.

For reading, I think I want mostly free to read. Can have subscribers only for reading stuff but I wasn’t planning to focus on that.

Not sure though. It’s awkward cuz a lot more people want to read than write. So how do you charge them if you give away reading? Maybe I need to make less stuff publicly readable. Not sure what or how to organize that etc.

BTW what’s good about subscription pricing? Limiting the amount of purchasing decisions people make. Don’t ask for their attention lots of times. E.g. if you had to pay individually for each post you posted, it’d be a hassle and distraction, even if the pricing worked out to the same amount for you. Similarly, I don’t wanna have a bunch of small individual products for people to buy because it takes too much attention for people to deal with.

Discussing with paying customers is generally higher quality. But it also excludes some people, especially young or foreign people, who would make good posters. Giving them free accounts on request won’t solve the problem b/c most of them won’t ask (the most important few may ask after reading the site a while and posting in the free section some, though). People will still be able to post on curi.us for free and on Discord (Discord will keep some free channels but might require being a subscriber to talk in most places, just like the forum might).

I’m inclined towards pay to post. I’ve tried free forums for ages. I think money generally makes things better. It helps keep the riff raff out, helps get people to take things seriously, and creates a “paying customer” relationship which has various benefits. But it’s hard to reach a conclusion b/c I’m not sure about the overall design b/c most ppl want to post little or none. What do I do with them? What would they pay for? Pay to read the majority of stuff is one option. That raises the barrier to entry to get involved. Maybe that’s OK if there is a free guided tour and if people like it and want more they should pay.

This is premised on the price being highly mainstream accessible, ($5-20/mo range – most ppl who might join, who can pay money at all, can pay that much). Not pricing like my $400 for YesNo. Having some high priced stuff is fine but I don’t want to make that any kind of requirement or general expectation; that’ll just be a minority of specific things as before. If people don’t care enough to spend a little money then maybe they shouldn’t be participating (applies to most people – to the ppl who could pay). I think that applies to reading a lot too. You probably shouldn’t be reading a bunch of unconventional ideas if you don’t care enough to pay a little. Even pay a medium amount. If you aren’t taking it that seriously, and won’t pay, then maybe you shouldn’t spend 100+ hours reading it as a lurker. Maybe more aggressive filtering will help people avoid bad experiences. If it cost $100/mo to read stuff beyond the introductory stuff, then a lot of people would have to make a choice about engaging more seriously or stopping pretending. And I kinda don’t think reading 100+ hours in a half assed way is a good idea. It’s not the worst idea. It has upsides and downsides. It’s kinda iffy. You’ll misunderstand a lot if you do that, and then if you start doing misunderstandings in your life you can make things worse.

There are issues with making it less accessible. Like I want to be open to public debate. I think asking for $5 a month if someone wants to debate me is fine. If they don’t care that much, whatever. (And I can make exceptions and give free accounts as convenient.) And that still applies for $10 and maybe $15 or $20. It doesn’t apply very well at $100/mo. And I do want people to be able to get started for free and preview it and stuff, but maybe that should be some specific things and just a small minority, rather than free read access to most of the forum. I do like being able to point people to public permalinks in general in debates. I also point people to books. If they won’t pay for a $10 book i’m just like “ok whatever you don’t care” (unless they have a good excuse like being age 10). Thread starters could maybe be allowed to tags posts “public” when they want to. One thing that helps is allowing people to sign up with no auto-renew. I think that’s customer-friendly. If someone wants to read a specific article I link and it’s a $5/mo subscription they may not want that, but if they are allowed to pay $5 to read it, as a one time price, and they also get a 30 day subscription but then their access will go inactive until they pay again, that’s harder to object to. There are also ways to do extra previews, e.g. view 3 articles per month free, which help with the permalink an article with an argument use case. Also I don’t expect to be debating non-members much.


curi at 12:50 PM on December 7, 2020 | #17 | reply | quote

#9 All those badges sounds like too much clutter, and also like they are about social status hierarchy. And kinda like gamification crap even tho the specific examples u used do have some value (like having a PF policy or learning plan post, but having customized code related to those things sounds overly special case to me. i'm more inclined to just let ppl have a bio/about/profile page and they can choose what to put in it).

Subscription level display on forum in general (instead of just profile page, or nothing) can be seen as a social status cue too. It (and some of the others) also can work as a sort of proxy for community involvement. I'm not even sure about name colors for subscription levels on regular forum thread pages. Maybe name colors for a formula to roughly indicate veteran status is better. Could be weighted factors based on e.g. topics started, comments posted, account age, and subscription level. Possibly a manual increase for people who predate the new community, too. I do find it useful, sometimes, to see some visual indication of whether someone is new or not.

Anyway, the subscription tiers would still need names on the purchase page even if they aren't displayed at all. Which is just a minor issue.


curi at 1:09 PM on December 7, 2020 | #18 | reply | quote

#10 Two separate companies, huh? idk much about that. The system sounds maybe dumb.


curi at 1:10 PM on December 7, 2020 | #19 | reply | quote

Subscription tiers could be named by the main features, something roughly like:

1. forum member

2. exclusive curi content

3. behind the scenes access

or

1. posters

2. bonus essays

3. group tutoring

(those names are poor marketing and the categories aren't decided but this gets the basic concept across)


curi at 1:20 PM on December 7, 2020 | #20 | reply | quote

A lot of trouble comes from social dynamics. What can be done to create a good atmosphere with better social expectations than the cultural defaults?


curi at 1:41 PM on December 7, 2020 | #21 | reply | quote

#17

> I’m inclined towards pay to post. I’ve tried free forums for ages. I think money generally makes things better. It helps keep the riff raff out, helps get people to take things seriously, and creates a “paying customer” relationship which has various benefits. But it’s hard to reach a conclusion b/c I’m not sure about the overall design b/c most ppl want to post little or none. What do I do with them? What would they pay for? Pay to read the majority of stuff is one option. That raises the barrier to entry to get involved. Maybe that’s OK if there is a free guided tour and if people like it and want more they should pay.

Another reason for pay to post is it would probably decrease abuse of anonymous/pseudonymous IDs.

I think finding enough paying customers for a philosophy site is a problem whether it's pay to post or pay to read or just pay to not be a moocher. Communicating a value proposition to people with little or bad philosophy knowledge (~everyone) is hard.

Do you have / can you create free content that will get a significant percentage of ~normal western people who encounter it from where they are to understanding a decent part of the value proposition of FI posting?

If you can, a monthly $5-$15 pay to post fee will be no more a barrier than the subscription fee is for Netflix. Millions happily pay it even though there are free entertainment options like torrenting.

If you can't, it will probably limit your forum to the people who already understand some of the FI posting value proposition, plus a few super-rare standouts who can figure it out on their own.


Andy Dufresne at 2:05 PM on December 7, 2020 | #22 | reply | quote

#19

> Two separate companies, huh? idk much about that. The system sounds maybe dumb.

I don't know which system you're referring to as maybe dumb: The legal system which lets people and companies be legally harassed to the point of destruction, or the idea of two companies as a defense. I suspect the latter but have low confidence in my guess.

If you don't annoy anyone much there's nothing to worry about.

If you only annoy people without significant money, power, and/or legal skills, there's probably still nothing to worry about unless you're big enough to attract ambulance chasers: lawyers looking for "victims" they can "help" by suing someone with deep pockets.

If you presume that no one will ever care enough to sue you then forming one company, let alone two, is a waste.

On the other hand, if you presume that you will someday annoy someone(s) with significant money, power, or legal skills - or grow big enough to attract ambulance chasers - then what? Do you have a plan to deal with that?


Andy Dufresne at 2:32 PM on December 7, 2020 | #23 | reply | quote

> I think finding enough paying customers for a philosophy site is a problem whether it's pay to post or pay to read or just pay to not be a moocher. Communicating a value proposition to people with little or bad philosophy knowledge (~everyone) is hard.

Making stuff mostly free, like now, doesn't actually fix the problem. If people don't see value, then people don't care, don't value it, don't put much effort in, don't share it much, etc.


curi at 2:33 PM on December 7, 2020 | #24 | reply | quote

#23 I meant it sounds like a possibly dumb system where "we registered 2 sister companies that aren't really independent, not one company, and split stuff up" makes a big difference for defending yourself.


curi at 2:35 PM on December 7, 2020 | #25 | reply | quote

#21 One thing that could help some with social dynamics is explicitly comparing FI to a quantum physics forum where some of the regulars are world leading quantum physicists.

People think they know a decent amount of good philosophy, and that ~everyone who has even a mild interest does. They don't think of it as a field like quantum physics where the can easily be out of their depth.

Or they think you're an amateur just like they are because you lack academic credentials and other prestige signifiers.

That leads to some social dynamics that aren't present as much when people see themselves in a learner role with someone who has world class knowledge in a complex field.


Andy Dufresne at 5:55 PM on December 7, 2020 | #26 | reply | quote

#24 Makes sense. Revised statement: I think finding enough customers who see significant value in posting to a philosophy site is a problem.


Andy Dufresne at 6:02 PM on December 7, 2020 | #27 | reply | quote

#26 That runs straight into social dynamics problems. People think it's *arrogant* to claim to be great at philosophy. Especially because 1) it's a field where they don't think there are good, objective ways to judge skill or merit 2) i don't have social status credentials like a phd, academic journal articles and endorsements from academic philosophers. Lots of people get really offended by the idea that I'm better than them at this, or I'm an expert, or they aren't very good yet. I'm not saying nothing can be done but it's tricky at best.


curi at 6:19 PM on December 7, 2020 | #28 | reply | quote

#25 If you actually operate as one company then the defense of nominally splitting things up doesn't work. It'll get tossed as a fraud (which it would be).

Just like if you create a company but then run it like an extension of your personal finances by not having separate bank accounts, or paying business bills personally and vice-versa, etc. In that case your company won't be much defense in a lawsuit; judge will rule you were acting personally rather than as a company and you won't have protection.

On the other hand if they're actually separate companies with different goals and business models, etc. are run as such and cooperate by reasonable formal agreements, the defense of having them separate is also pretty reasonable.

A company that builds forum software and licenses it to people wanting to host forums is in a different business from a company that hosts a particular forum and charges forum users for subscriptions. They can easily be separate companies in fact, not just in name. And as such if you sue one into bankruptcy, you don't get either control of or assets of the other one.


Andy Dufresne at 6:19 PM on December 7, 2020 | #29 | reply | quote

#29 oic re has to be a real split


curi at 6:22 PM on December 7, 2020 | #30 | reply | quote

#28 Ya I see the arrogance problem.

It seems to me that to get anywhere people have to:

- Have a very unconventional (for adults) willingness to adopt a mostly conventional learner role despite your lack of credentials

or

- Have very unconventional attitudes and skills around comprehending, productively discussing, and changing their mind about material they think is obviously wrong.

Are there other ways we know of that have a decent chance of working?


Andy Dufresne at 6:22 AM on December 8, 2020 | #31 | reply | quote

social expectations

#21

> A lot of trouble comes from social dynamics. What can be done to create a good atmosphere with better social expectations than the cultural defaults?

Have something in writing, describing in what ways the social expectations here are different from cultural default social expectations.

What are the ways that social expectations are different here? I don’t think I could accurately say.


Anne B at 8:02 AM on December 8, 2020 | #32 | reply | quote

#27 The site emphasis doesn't have to be on posting. Posting is a good feature to have. But I think the majority of new people shouldn't start with posting. They don't know how. They aren't ready for it. It's overreaching.

And "how to post" is not the first thing anyone should be learning, either. Maybe the 10th.

What tends to happen when people try to learn stuff without discussion is:

- misunderstand stuff & don't get corrective feedback on misunderstandings

- don't integrate it into their life (even parts they understand OK)

- don't practice. learn things to "conscious competence" at best, not to mastery/autopilot/"unconscious competence".

- lack of initiative. go through a few available resources and just kinda stop doing much instead of vigorously pursuing more leads to learn more things.

The area people should work on first is about how learning/thinking/reason work. That underlies dealing with all other topics including what to do in discussion.

More of a "guided tour" learning experience, plus worksheets and answer keys, would help with some of this.

Weekly Q&A stream would help some.

Weekly new article (and/or slide-based video) on CF would help keep people focused on the right areas (how to think; how learning works) instead of e.g. politics.

Not everyone is going to bring much attention, initiative or creativity. Things can be made easier for people but not easy by the standards of the majority. That's OK. 0.1% of just Americans is ~330k people.


curi at 10:51 AM on December 8, 2020 | #33 | reply | quote

#31 They can also join the community and gradually start learning its norms and values, making some friends, etc. This would be significantly easier if the community was a bit larger and most people did *not* have a lot of critical contact with me early on (too hard for them if I try to treat them like a a competent, serious person – a peer/equal besides background knowledge).

They can also learn something related (particularly CR, TOC, or Oism) and then use that as a head start on FI. This sometimes runs into conflict b/c they think they're much more competent at that prior knowledge than they are.

They can learn privately so the learner role and social status aren't as much of an issue. A lot of people try to do this to some extent but I think need a more complete and guided experience.


curi at 10:59 AM on December 8, 2020 | #34 | reply | quote

>> A lot of trouble comes from social dynamics. What can be done to create a good atmosphere with better social expectations than the cultural defaults?

> Have something in writing, describing in what ways the social expectations here are different from cultural default social expectations.

> What are the ways that social expectations are different here? I don’t think I could accurately say.

People are bad at consciously recognizing how social stuff works or reading something about social and knowing how it'd apply in life/practice/action. You can't just tell people the answer and expect it to work. They have to practice, create new and better intuitions, etc. And before that they better know some philosophy so they have an understanding of what they're doing and why (like what practice, mastery and intuition/autopilot are). And they better be pretty invested into this, and care a lot, before trying to change a lot socially.

Having a few comments in writing can help in some limited ways but it isn't going to fix this and the comments shouldn't be expected to be complete.

Most of what needs to be done is stuff like setting the tone, providing good examples/role models, presenting things in a way where doing X feels more natural to people not Y, designing things to encourage good behavior and avoid unnecessary heavy clashes with existing norms.


curi at 11:20 AM on December 8, 2020 | #35 | reply | quote

I think I like my four subforum idea: main/details/meta/other. So the main one is reasonably accessible for most users and works well for them to follow. Advanced philosophy could go in details (aka technical) if it requires significant background knowledge – that isn’t just for math and coding.

Feedback on having those four main categories/subforums? Can anyone think of posts that wouldn’t fit into any of those four well? Anyone doubt the point or think they wouldn’t like browsing with that system? Anyone think people would get stuck on figuring out which category to use? Any other categories you think are important or useful?

Other sorts of tagging that can be value are topic tags. Also: debate, discussion, tree creation, article, long term project, serious project, question, unbounded project, wanting unbounded criticism, wanting tangential or framework criticism. High or low confidence in what you’re saying, or speculative ideas vs. developed ideas or exploratory writing. Some stuff like that. Also there’s the side/margin note vs. main point distinction. I think that could be useful in some threads but probably as a standard thing people bother with all the time. Also marking posts as decisively refuted (and by which other post) and other idea idea tree related metadata (e.g. mark something as agreed on by the discussion participants according to me – this should probably be on a per user basis (e.g. i can mark it the way i think it is, but someone else could disagree and attach different metadata). Another type of tree meta data is grouping: i think this group/subtree of posts work together to make an argument and should, in some sense, be treated as an integrated unit.


curi at 11:50 AM on December 8, 2020 | #36 | reply | quote

Long term big picture goals: I'd like "spend a few years active on this site" to be more effective for learning than high school or university. I'd like to give people an alternative or supplement to shitty schools. (It did that for me, btw. I quit university and learned with DD + the community. But I'd like to do helping in a one-to-many way not primarily one-to-one.)

FI already has some advantages over the schools and getting a better outcome here can happen for much cheaper. But FI also has some notable risks and downsides. It's much less structured and guided. People need a lot of initiative to do well here. That can be reduced some. FI is also risky to social hierarchy climbing strategies, which could also be mitigated somewhat without loss of integrity, though there's also a major clash of principles there. If FI could be reasonably accessible to 5% of people instead of 0.01% that'd be great.

I think maybe my grammar article and CF course could be pretty helpful to more than 1% of people. I think I can make some more accessible educational material if I focus on doing it. I also can help guide people more and debate them less. My debates with people have partly been about a clash between 1) my expectations about the world 2) the world as it is. Having a more accurate view of what the world is like, what exists and doesn't exist, etc., allows me to get along with people better and pursue projects that will actually work. (There are some difficulties there, e.g. people who want to debate me as a peer unless and until they recognize me as having very high social status. One of the ways to deal with this is better, besides opting out, is structuring debates to make outcomes clearer.)

I have other goals too, e.g. improving the state of the art in epistemology and sharing that. The goals are related and complementary.


curi at 1:14 PM on December 26, 2020 | #37 | reply | quote

#36

> My debates with people have partly been about a clash between 1) my expectations about the world 2) the world as it is.

This is hard because of dishonesty. People say they want criticism who don't. People say "this is a think tank full of really smart people doing great work and having really rational discussions" when it's not. And a lot of dishonesty is mutual, e.g. Joe says he wants criticism, and Jane says she's criticizing his work, but the criticism is very limited or non-existent. Or there are people third parties who will say an irrational think tank is a wonderful beacon of rationality, and that's not just because they get fooled, it's also because they are complicit some and, related to being complicit, they have low standard for what words like "rational" mean and how to communicate about stuff (it's kinda like they exaggerate a lot, to the point of dishonesty, and bake that into their thinking at a low level).

And if you call out the dishonesty, that *also* leads to clashes with the world, just like partially believing their lies does.


curi at 4:29 PM on December 26, 2020 | #38 | reply | quote

> I think I like my four subforum idea: main/details/meta/other.

mb change 'details' to 'detailed'? I think it fits "that isn’t just for math and coding" better b/c it's an adjective. like with 'details', do I start something elsewhere and move it/start a thread in 'details' when talking about finer details but not broad stuff?

Some other thoughts: the (main, details) pair feels like it has a special relationship compared to 'meta' and 'other'. some other name pairs that sound similar to me: (general, advanced), (main, aux), (basic, advanced).

> Can anyone think of posts that wouldn’t fit into any of those four well?

Not atm. but I am wondering a bit about moving topics between them or transitioning between them as conversations progress.


Max at 8:20 PM on December 26, 2020 | #39 | reply | quote

The simplest and lowest barrier to entry way to have paying customers is a *one time fee for lifetime forum access*. So that has some advantages over a paid subscription as the bottom end pricing.

The one time fee is a major barrier to entry if it's large, e.g. $1000 for lifetime access to read and post at the forum. But if you do $20 lifetime access instead of $10/mo, then it's a smaller barrier to entry and it's easier to sign people up.

So that's worth some consideration. $20 one time is enough to do some filtering of people who don't care and to break the ice on paying any money at all. Higher prices do filter out unserious people better. And people who feel entitled because they paid a low amount can be a problem and a customer support burden.

The downsides are 1) get paid less (or have to upsell them other stuff later) 2) i don't think price is the barrier to entry for most people (if they care much, and they are the right kind of person for the forum, most of them could pay $20/mo. the ones who can't are not going to buy much anyway, though i'd still prefer to have them. but of the ones who can't, a decent number can't do $1/mo either, e.g. they are a middle school kid who can't buy stuff online without asking his parents, so the problem for them is paid not free, rather than the amount of money. **broadly a LOT of people are not very price sensitive once you get them to actually pay, and if you set prices low you give up a lot of money from most of your customers.**).

the question of how low a low end pricing you want is relevant for monthly subscriptions too. could be $5/mo or $25/mo as the lowest price, or in between, or even more. i wasn't planning more because i figured tiers would address that. my current thinking is roughly:

1 month free trial (btw if they post a bunch of good stuff and say they can't pay, we could just give them a free 3 month extension or something)

tier 1: read and write at forum

tier 2: one-to-many help, e.g. can join my 2 group classes per month on zoom, and access to my topic of the week thread/article

tier 3: behind the scenes stuff (meta stuff e.g. participating in a planning thread like this one, as well as "how it's made" type stuff like early exploratory article drafts) and maybe advanced stuff, and maybe some personalized attention

i think these tiers are problematic because if someone signs up i want them to have good experience and learn stuff, but i'm pushing some of the help they should have into tier 2. tier 1 is kinda a "do it yourself" tier which is unsuitable for most ppl so they should be steered away from it.

also i think some planning stuff is reasonably general interest, and safe to share, and should be visible to everyone. let ppl be involved and give feedback.

a different way to view tiers is:

1: lurker tier

2: actively trying to learn tier

3: not price sensitive or superfan tier

But if you exclude lurkers from some learning activities then it's harder for them to upgrade and get involved. But OTOH i want an offering that makes sense to ppl who aren't gonna participate much.

There could be a tier where you pay to read and then tier 2 allows posting, but that makes it even harder for borderline ppl to start getting more involved, so i think that'd be bad. And I think if ppl are paying at all they should be allowed to post.

Some ppl make forums more secondary btw. https://thefutur.com/pro-group they have a youtube and sell courses. forum is 150/mo and has > 300 members. the marketing is focused towards newer ppl and says you'll get to talk with ppl who know more than you and are making more money and you get a safe, private space to answer questions and get some help from experienced ppl.

Jonathan Stark is 150/mo for 2/mo group coaching sessions and it comes with a Slack chat forum as an extra bonus. "Group Coaching also includes a dedicated Slack channel to facilitate discussion in the two weeks between live video sessions." https://jonathanstark.com/group

Binswanger focuses on a 12.50/mo forum subscription:

https://www.hbletter.com

> Launched in 1998 as an email list, HBL now offers: weekly Zoom conferences, emailed posts by Harry Binswanger and selected others, a public blog, a private Facebook group, and a discussion forum for subscribers to HBL ($12.50 per month).

i think pricing is unclear for me due to lack of focus. i want to do multiple things:

- let lurkers hang out and read stuff (maybe that should just be the free tier. take a few selected things and make them public on a regular basis. that might make a free trial unnecessary too).

- have a DIY forum where members chat with each other, try to learn, etc.

- have a more involved "i am trying to learn" group that gets a few things and a bit more help

- have a "ppl who care a ton" group that gets some more stuff

i'd like to do something like "$500/mo. learn stuff. better than college". but i think it requires having > 10 ppl signed up or it's too much work for me. and that kinda thing might work better as discrete courses instead of a general monthly subscription.

i think a price point like 100/mo or 150/mo for a forum excludes too many ppl as a bottom price. i'd like something more accessible than that. but pricing like that is reasonable for e.g. group tutoring.

*maybe public/private is an important distinction to charge for*. i was envisioning the main forum as something ppl should treat as public even if something is paywalled. there could be a private section at a higher price which ppl are expected to not share. like:

$12/mo DIY (more stuff to read than the free material, can discuss)

$50/mo some group help, some structure, guidance, scheduled activities

$150/mo private forum section (not like NDA level secrecy, just ask everyone not to share stuff elsewhere and ban ppl caught leaking)

10% discount for annual billing.

a private forum also fits somewhat with my "behind the scenes" tier concept. i could post some stuff there which is not of general interest so idc about public discussion of it. (and i can always repost/reuse my own stuff later if i want, i'd just lose all the replies if making it public.)

what's the point of the private forum when there is a public one too? why do ppl want privacy? they are worried about asking dumb questions or looking dumb publicly. they want to do some of their learning in a secret, safe way. the private forum could also be less critical and have a different atmosphere (tho even for the public forum i plan to make debate or desire for criticism more explicit and opt-in before providing major crits).

would really have to emphasize 50/mo as the main tier if you're expecting to learn. not sure about that.

a different tier idea is "i want ~unbounded criticism from ET" (rly less bounded) and you have to pay to opt in to that.

my 500/mo "better than college" tier idea would have at least one class/wk, prepared slides ... oh i see the problem. ppl are gonna be at different places in their learning b/c some have been doing it for a year and some are new. so some of them basically need the material i did last year and some don't. so organizing as separate courses by topic handles that way better than having everyone subscribe to the same ongoing thing. my initial thought was basically if 20 ppl signed up today i could do the same activities with all of them and it'd work fine. cuz basically no1 has enough background knowledge that they need to skip stuff. but once i teach some stuff then new ppl join after that, i'll have students at significantly different places in their learning. so yeah nvm that, individual courses by topic are better.


curi at 12:56 PM on December 27, 2020 | #40 | reply | quote

Pricing thoughts

#40

Do you want to make a lot of courses over time? Do you want to do interactive ones like the one in Nov or non-interactive like YesNo?

If you want to make a bunch of courses I think a couple additions make sense:

- Each course should have a specific forum area or parent topic for discussing the course content. Only people who have access to the course should have access (read and post) to that course's area / topic. If they buy the course individually, they should have lifetime access to that particular area / topic even if they stop paying for other FI stuff.

- In addition to selling individually, offer a "content all access pass" tier or add-on option to the top tier that gets access to everything. If you're doing interactive courses I'd imagine the pass would get recording access but not interactive enrollment. And the access isn't lifetime, only for as long as they continue paying for the pass.

There's a risk someone joins the content tier, downloads all the content, then cancels. You could address that with DRM or with contract periods but I think both would be bad. Even without DRM or contracts they'd lose access to the forum area / topic for the courses and any new material you create. I'd look at it like a kind of refund request. Is somebody who signs up for 1 month and then cancels really going to get the same value as somebody who legitimately subscribes to & discusses the material? I don't think so.

It may also make sense to do grandfathered pricing for the access pass. Ex:

For 2021 you set the all access pass at $200/mo given your current content and what you expect to create over the year.

For 2022 you set the pass at $250/mo because of the additional value of content created in 2021. But people who signed in 2021 continue to pay $200/mo.

For 2023 it's $300/mo, etc. but people continue to pay the price they signed up at as long as they stay signed up. If they cancel and come back, they have to pay the new price.

This makes sense because the people who have stayed longer have paid more over time, it provides an incentive to sign up early and stay signed up.


Andy Dufresne at 6:47 AM on December 28, 2020 | #41 | reply | quote

Criticism

#40

> a different tier idea is "i want ~unbounded criticism from ET" (rly less bounded) and you have to pay to opt in to that.

I think this would be good with two additions:

1. The criticism is private rather than public or semi-private.

2. (more elaboration on "less bounded") The criticism is targeted towards helping with the problem(s) the customer wants to solve.

I think those are the two values ppl would be most willing to pay for.


Andy Dufresne at 6:57 AM on December 28, 2020 | #42 | reply | quote

> Do you want to make a lot of courses over time?

Possibly, but not regularly without more students, so a subscription wouldn't work well currently.

> Do you want to do interactive ones like the one in Nov or non-interactive like YesNo?

Both.

I'm familiar with grandfathering pricing and generally agree. I view it as the default. I'd need a clear reason to do something else.

Yes subforums for each course make sense.

> 1. The criticism is private rather than public or semi-private.

Fully private one-on-one criticism is available at my consulting/tutoring/coaching rates.

> 2. (more elaboration on "less bounded") The criticism is targeted towards helping with the problem(s) the customer wants to solve.

It's easy to try to do that and offend people by questioning their goals and their available resources for the project, particularly background knowledge. Those are two big ways the discussion can get away from the topic they had in mind, but they're often highly important and relevant.


curi at 11:53 AM on December 28, 2020 | #43 | reply | quote

The cheap tier could be rate limited, e.g. 10 posts/month. But I think that's a bad idea, even though part of the point of a cheap tier is for lurkers. But only the lurkers who care a bit more and want more to read, or else people could just read free stuff now and then.

Pricing depends partly on marketing. How many people are we finding? What type of people? Is signing one up at $12/mo a good result or somewhat of a waste of a high quality lead?

Do I want cheap users at all? I think so. I'd like to have a pretty accessible forum. I'd like a bunch of people in the community rather than a small, exclusive group. I think I'd like a higher volume forum where most people make choices about what to read, rather than a small forum where it's common to read everything.

Can get more people for free but I think that's problematic for discussion quality. Though I was thinking maybe one free open thread per year to let free users talk at all (maybe requires account but no payment info), e.g. ask pre-sales questions, or they could discuss with each other. People who want to talk for free could also go post on curi.us or FIGG, which I wouldn't encourage but could just leave alone.

Idea: could have family plan discounts.


curi at 12:39 PM on December 28, 2020 | #44 | reply | quote

#4

> - how much should stuff be personalized and 1-on-1 attention type stuff vs. impersonalized digital course and tutorials type stuff?

I wonder if this kind of thing would work: You (curi) pick a topic and provide learning materials (stuff to read or watch or listen to, exercises) and a schedule/structure. People taking the course work on their own on learning each section. Then they get together (could be synchronously or asynchronously) and discuss and compare what they’ve done.

This would require work from you on the front end, but not ongoing attention. From your point of view, it would be impersonalized. But the learners would get personalized feedback from and discussion with other people who were thinking about the same ideas at the same time.

It would be like your grammar essay but with more organization. One thing I liked about that grammar essay was working on my own to learn the stuff and do the exercises, and then being able to see other people do the same exercises and to discuss the ideas with them.

I can think of some potential problems with this kind of course:

If none of the students understand the material very well, they can’t help each other much.

Students will come into the course with different starting knowledge/skills.

If enough students don’t do the work in time to fit the schedule, the discussion part might not be valuable.

It would be hard for the course author to get feedback on the course if he’s not there to see how it goes.

Now I’m thinking about how I’d turn your grammar essay into this kind of course… Actually, in my mind I’d be involved in the group discussion/feedback to see how it was going and to help if needed. So I don’t know.


Anne B at 9:04 AM on January 6, 2021 | #50 | reply | quote

Pricing Thoughts

My current thoughts. Feedback?

Free:

Some public stuff that people can start learning from

curated selections from forum

marketing materials

can talk at one public open thread (requires free account)

$30/mo:

Read my stuff (weekly article; ~daily small posts)

Watch my stuff (e.g. regular videos, streaming, group classes/Q&A; unlisted on YT)

Read the forum

Discuss on the forum

Structured learning activities (e.g. topic of the week, book of month, worksheets)

Ask questions

$50/mo:

~private forum section

$150/mo:

behind scenes stuff (some planning, moderation, meta, community stuff. some marketing and biz discussions. some exploratory writing)

Other:

Tips/Donations (also can sign up for any subscription amount, not just the tiers)

Digital goods

Tutoring, coaching, consulting, advice, writing for hire

Live courses

I don’t want to sell cheap/small things. I don’t want to ask people to make lots of purchasing decisions. Just have the subscription pricing and some big, high effort products. If you pay money, you get something substantive. If it’s too small to sell, I’ll just give it to subscription members.

I don’t want to use advertising for revenue.


curi at 12:55 PM on January 12, 2021 | #51 | reply | quote

an idea for the site name: criticism.gift

FYI i bought that domain just now and am willing to donate it if we want to use the domain.


Max at 8:30 PM on January 14, 2021 | #52 | reply | quote

problem with 'forking' threads?

I think there might be a problem with someone creating a new top-level post that is also a comment/reply.

What happens with replies to the new post? My intuition is that different things should happen for different threads.

If we take FIGG norms as an example: when you fork a topic (reply all & new subject + (was: X)), filtering by the original topic won't bring up the new topic + replies to the new topic, typically. (Google groups UI seems to do this, but also hides the new thread as a consequence).

My intuition is that we want to include all replies when the new and previous topics are closely related. we don't want to include all replies when the topics aren't closely related (which mb includes topics started to discuss meta).

Mb an alternate soln for closely related topics is like creating a subtopic within the thread. For a thread view, we could show a list of the titles of subtopics near the top, for example. then provide easy filter access so that you just look at replies to the subtopics.

For like forking a topic, that's more like FIGG where the new topic and replies aren't included in the original topic. Then the behaviour is consistent, and it better matches my intuition about how it should work.


Max at 9:21 PM on January 14, 2021 | #53 | reply | quote

there are two types of forks:

1. rename topic

2. move topic

moving involves linking to and/or quoting from a node, but giving your post a different parent, e.g. the Main node.

renaming involves setting a title on your post.

it'll be a lot easier to understand after having the core features working so you can try them out. it's harder to plan ahead about forking now, and unnecessary. rapid prototyping and iterating is the way to go instead of trying to imagine everything in advance.

the core features are things like creating a node, specifying a parent, quoting stuff, linking stuff. those are the ones that should be considered before coding, but not details like thread forks.

some design considerations re the core features have been discussed above, like:

- post graph or tree?

- two types of connections between nodes? parent/child (main tree) and reference (graph with cycles). two separate graphs?

- two types of node: regular post or side comment

these are the kinds of decisions that are harder to change later.


curi at 9:41 PM on January 14, 2021 | #54 | reply | quote

nodes will have various properties, e.g. probably a list of tags.

i think they should have a default_view property or something to accomplish a similar goal. some way to tell the system to display like a forum index, a blog post with replies, a blog comment, or possibly something else. nodes like Main and Meta will use the forum index default, and by default their children will use blog post and their grandchildren use blog comment (in other words, if the property is blank, traverse up the graph to figure out the right default).

this would let you create a subforum, e.g. "politics", anywhere, e.g. as the child of Other or 50 levels deep in a discussion.

an issue to consider is the usage distinction between subforums and category tags.

another consideration is how to display mixed pages. my first guess is the OP (root of the subtree currently being displayed – basically the way you view the forum is point your browser at a particular node and see it plus some descendants) should control the display type and the display settings of everything else on the page don't matter.


curi at 9:48 PM on January 14, 2021 | #55 | reply | quote

#55 How does this look as a starting point?

!()[https://user-images.githubusercontent.com/1046448/104728758-84803f00-578b-11eb-944a-6cf0af2e1696.png]

the tags are denoted like: `value :: type` and blog_c::view means the blog_comment view


Max at 4:45 AM on January 15, 2021 | #56 | reply | quote

Fixing image:


Max at 4:46 AM on January 15, 2021 | #57 | reply | quote

#51 the middle tier with private forum can also include chatroom access (including both semi-private and quoting-in-public-allowed rooms). discord, slack or other.


curi at 12:11 PM on January 15, 2021 | #58 | reply | quote

I have a very basic partial prototype at https://cff.au.ngrok.io/.

Nested replies example: https://cff.au.ngrok.io/6

If you post a reply then you'll get a page like: https://cff.au.ngrok.io/10 -- click on 'to parent' to traverse back up to parent comments/post/etc

The root is at https://cff.au.ngrok.io/0 (or you can omit the `0`)

I'll try to leave this page accessible if anyone wants to check it out.

code: https://github.com/fi-tools/cf-forum


Max at 7:53 AM on January 16, 2021 | #59 | reply | quote

I still like these 3 tiers:

1: member (includes all learning activities, e.g. group tutoring streams)

2: private subforum & chatroom

3: backroom

pricing ballpark i think 30/50/150 is ok. i'm thinking do 3 or 4 month subscriptions b/c i think doing year-long lock ins is kinda sketchy. i want customers who want to be here. but i'd like ppl to sign up for a significant trial period, like 3 or 4 months, not 1 month, to try to get them to give it more of a try and think of it as a longer term project. 12 month projects seem too long but 1 month is too short for ppl who are pretty busy and are just doing it on the side, not focusing on it full time. (can also allow 1 month subscriptions at a higher price per month.)


curi at 1:43 PM on January 18, 2021 | #60 | reply | quote

Idea to filter out bad people: block posting from phones.

Less harsh idea: put a small label or color on posts from phones, and allow filtering them out.

On average, phone users are lower quality and dual-use people write worse stuff when on phones.

smartphones are a big part of the mainstreaming of using the internet but the mainstream is dumber than the early or more serious users.


curi at 1:18 AM on February 13, 2021 | #61 | reply | quote

#61 maybe start with just paywall, plus track mobile posting internally, then consider more screening if there's a problem. that has some issues, e.g. some bad cultural norms could get started and might have to give out refunds and alienate some ppl if i let them post from mobile then took it away. changing from internal tracking to publicly visible labels of mobile posts would be more minor but there isn't a lot of downside for starting with that and having a culture that discourages posting much on mobile.

mobile posting is fine for side comments or whatever we call the secondary type of messages. btw Basecamp kinda has that as "Boosts" and limits them to 16 characters. you can just do an emoji like a discord react or use it like a twitter like if you do a heart or thumbs up. u can also type a couple words. it's more flexible and seems better to me than emoji-only or a like button. my initial plan was to have side comments with a visual indicator and ways to filter them out or collapse them, but they allow full features. you can use them to say "+1" or "thanks" or an emoji and you won't clutter the main discussion. but i'd expect 1-3 sentences to be common, including saying afk until 2moro and meta questions like "Have you read BoI?"


curi at 11:12 AM on February 13, 2021 | #62 | reply | quote

The new website is gonna take a while. I decided to try using Basecamp in the meantime:

https://3.basecamp.com/4983193/join/KEHzLZV59MYT

I think it's a lot better than Discord and I have been unable to get ppl to just use curi.us. I still prefer curi.us in general but Basecamp does have some nice aspects that I'd like to get more experience with and see how people respond to (maybe we can copy a few ideas for the new website). The project management design isn't a perfect fit for a public discussion group but *people should treat learning more like an organized project* and take it seriously, so there are upsides too. And Basecamp is a pretty simple, flexible, freeform tool, so it can be used OK outside a business colleagues context.


curi at 11:19 AM on February 15, 2021 | #63 | reply | quote

#63 Basecamp doesn't support markdown and doesn't have a good ways to import or export writing. Those are my biggest concerns. Basecamp also doesn't do public permalinks. People have to join to read anything.

If those things concern you for a particular thing you write, put it elsewhere (e.g. here or your blog), and you can copy the text over to basecamp or share a link. If Basecamp is a bad fit for a particular discussion, just do it here. I think the main solution to Basecamp's limitations is: don't try to use it for everything.


curi at 11:21 AM on February 15, 2021 | #64 | reply | quote

The new Basecamp ( join here ) is for self-directed learning with some tips and tools.

I want the new community website to require less self-direction. I want to take the lead more and make it easier for people. People don't know how to self-direct or self-manage progress or learning, and need help getting started before they do that. (I'm available for hire for anyone who wants help now.) I want to figure out how to mass-produce help so lots of people can improve (at an easily affordable, mainstream price point, rather than the higher price of personal help just for you).


curi at 12:56 PM on February 16, 2021 | #65 | reply | quote

I want the new site to work even if people don't participate. They can be quiet and I'll just keep posting good, helpful stuff regularly.

I do *not* want to do that with the Basecamp. That's a free thing for people who actually want to do stuff. But people mostly won't do stuff (e.g. no one has done 5 tiny projects with planning or, possibly even easier, retroactive planning, which could easily be done in one day). I think they need to do time tracking first but most of them either won't do that or won't share about what's going on.

A paid thing that's easier to use is good.

It's hard to set expectations on other stuff like a free forum b/c ppl mostly don't listen. (I did a post about expectations. No one replied. No one tried to discuss it until they actually understood what it meant.)


curi at 1:50 PM on March 2, 2021 | #66 | reply | quote

goals:

- paid membership

- i share polished stuff

- i share casual stuff, e.g. links, microblogging

- other people can post/discuss

- organized well to enable discussion

- tree or graph

- notifications system

- no destructive edits or deletes of posts (versioning system)

- search and filters

- multiple, useful views of content

- markdown

- permalinks

- main comments and side comments

- organized well to make it easy to follow my polished stuff or all my stuff

- participation optional. i don’t ask or expect ppl to talk, practice or learn. can be a nice place regardless of what they do

- i’m more aloof and focused on writing my content. avoid much parochial discussion.

- more barrier to entry to get criticism from me. rly make sure it’s wanted and asked for in that case.

- atmosphere where ppl can be wrong and i ignore, even if they use my name and key terms. no expectation that i’m checking stuff for errors.


curi at 3:07 PM on March 11, 2021 | #67 | reply | quote

#67 the items from "tree or graph" to "main comments and side comments" are supposed to be indented one level.


curi at 3:08 PM on March 11, 2021 | #68 | reply | quote

another goal i forgot on list:

- free tier is curated content w/ no discussion


curi at 4:10 PM on March 11, 2021 | #69 | reply | quote

seeing some drafts and editing process (my own and with feedback from others) could work well as a tier 3 perk, a thing to do in the backroom area.


curi at 3:24 PM on March 24, 2021 | #70 | reply | quote

Probably going to go with forum software that someone else already made. Taking recommendations if you know something you think would be particularly good. Discourse is the leading candidate atm so if you can say "X is better than Discourse at Y" that could be useful.

Currently planning to make posts publicly viewable but charge a one-time fee to be able to post. The point of the fee is to raise quality and reduce harassment, not to make money. Probably try $20 and raise it if there are problems.


curi at 10:58 AM on April 18, 2021 | #71 | reply | quote

Want to discuss this? Join my forum.

(Due to multi-year, sustained harassment from David Deutsch and his fans, commenting here requires an account. Accounts are not publicly available. Discussion info.)