Specialist Creators with Small Audiences

Elliot Temple
,
41

Those who do not commit fraud themselves usually tolerate it in their peers. The minority who will not tolerate frauds usually weed themselves out quietly. I have lost count of how many friends of friends entered a PhD program, had an adviser who tacitly or explicitly demanded they commit fraud to get publishable results, and quit in disgust without raising a public stink.

A Plan to Improve the World: 20 Rational Debate Advocates
42

I listened to this companion podcast to that article some time ago:

I find Samo Burja and Bismarck Analysis interesting. Iā€™ll read that article.

43

From the podcast (typed by ear by me):

Burja: By the late 19th century it was rare for someone to be doing science outside of a universities uh context.
Landau-Taylor: Right, and like Einstein was. And the universities could still get the science from him, but that was like unusual and surprising in his day. And today I donā€™t think they could accept the contributions from an Einstein who was outside of the system at all.
Burja: So if a patent clerk today actually came up with better theory of gravity, they probably would not recognize it.
Landau-Taylor: Yeah, like theyā€™d publish on weird blogs and like after four years some guy with a phd would rewrite the thing and then they could take it seriously.

So thatā€™s the Man-German story.

Then he brings up Guzeyā€™s refutation of Why We Sleep as an example.

44

Woah, this post made me tear up.

1 Like
45

i was going to post in https://discuss.criticalfallibilism.com/t/curiosity-specialist-creators-with-small-audiences but i canā€™t find it now. guess it was deleted because the auto generator mistakenly created an unnecessary copy? i havenā€™t read through this thread yet. i was initially looking to post this in the i changed my mind thread but this fits better here after reading the article.

i think LMD was going in the right direction in the historicism thread getting to the same issue that iā€™m raising there but then stopped. asked three questions about your article, no reply, never posted in that thread again. you just acknowledged in #21 that there are multiple criteria and the analysis isnā€™t as clear-cut as the original article presented. LMD could have got you there months ago.

honest observation from my experience so far: itā€™s harder to have a productive back and forth with you than it usually is with people i discuss ideas with or even just people in general. thereā€™s something about how the conversations go thatā€™s hard to explain well. i think it reflects in how quiet the forum is unless itā€™s been unusually quiet since i joined. maybe itā€™s just a bandwidth issue. i donā€™t know if creating for an audience is a full time thing for you.

i pointed this out in the capitalism forum thread too that i almost stopped because your reply felt closed. i believe you when you say you didnā€™t mean it that way but itā€™s definitely an issue for this forum to work as an error correction system.

your article frames the problem as people not caring enough but what if some of them cared and just stopped trying? saying you outclass people at debate reinforces this. it tells the copilots theyā€™re right to stay quiet.

46

Yeah I got more of the old threads properly linked to the blog posts today. A bunch werenā€™t which leads to duplicates.

1 Like
47

I donā€™t think I acknowledged that. Please be careful speaking for other people.

1 Like
48

acknowledged was the wrong word. i was trying to say that in #21 you describe historicism as having multiple themes rather than one clean criterion, and you say some of Deutschā€™s claims look non-historicist on the surface. that made the issue seem less clear-cut to me than the articleā€™s framing. is that a fairer way to put it?

49

you just acknowledged in #21 that there are multiple criteria and the analysis isnā€™t as clear-cut as the original article presented.

(I didnā€™t acknowledge that. Instead of claiming I made a mistake, this presents your criticism as my own opinion, which is unfriendly and negative.)

LMD could have got you there months ago.

(This accuses me of irrationality and hypocrisy. Itā€™s saying I resist error correction and stay wrong. But itā€™s begging the question. We didnā€™t actually start, let alone finish, a debate which concluded that I was wrong. This follows on from the previous part which falsely claimed that I had already conceded, in which case this wouldnā€™t beg the question.)

i think it reflects in how quiet the forum is

maybe itā€™s just a bandwidth issue. i donā€™t know if creating for an audience is a full time thing for you.

(Implying that I donā€™t respond enough and/or my posts are too low effort.)

i pointed this out in the capitalism forum thread too that i almost stopped because your reply felt closed. i believe you when you say you didnā€™t mean it that way but itā€™s definitely an issue for this forum to work as an error correction system.

your article frames the problem as people not caring enough but what if some of them cared and just stopped trying? saying you outclass people at debate reinforces this. it tells the copilots theyā€™re right to stay quiet.

This is a lot of negativity without the specifics needed for me to effectively make changes or debate your claims. The claims were not established in prior discussions that reached conclusions. You have not yet requested to debate anything (yet seem to have already concluded that Iā€™m wrong about some things). And itā€™s too many complaints to address simultaneously. Iā€™m surprised because I thought you had given positive feedback. For example, you wrote in AGI, LLMs, CR and CF :

oh right, my bad.

i think i buy it for now.

Ok the binary evaluation makes sense to me.

I read your intuition articles and it makes sense

iā€™m confused about something

ok cool, that clarifies a lot.

Your point about intellectual tolerance and humility brings up an interesting AI design question.

cool, that clears it up.

one of the most fascinating ideas I came across was on sam harris podcast it would take a lot of work to understand it deeply.

Then I explained it to you and you said:

oh ok thanks.

Thatā€™s only from the first 30 posts in one topic. Another example was when you Liked this post of mine:

I can often write short, curt posts that arenā€™t warm and friendly, or I can write fewer posts. Social dynamics can cause them to be taken as unfriendly, but theyā€™re actually me being friendly and replying more than most public intellectuals would have.

Youā€™re welcome to post a lot and ask a lot of questions as long as I can write short replies, and not reply to other things, without it being taken negatively. These things go together :) (Bringing up the same topic again that didnā€™t get an answer is fine btw, though itā€™s usually best to change something, e.g. doing more of your own analysis or asking a different question about it.)

Now youā€™ve raised this issue again without responding to the answer I already gave that I thought you had already accepted.

Also, if you know of any similar forum which is at least half as good as this one, or another philosopher who is open to debate, please share.

50

where?

51

You can look up stuff in a particular post by writing it in quotes and selecting ā€œsearching in this topicā€:

image
image1016Ɨ274 26 KB

It took me just a few minutes to check. Hereā€™s the link to all the stuff Elliot quoted in the order he shared them:

** I think @Elliot may have made a mistake with this one. @ActiveMind had wrote it, but @Oracle did heart/like the post (maybe thatā€™s why he included it?).

2 Likes
52

Thanks, I edited to fix that.

1 Like
53

@Eternity thanks for checking. and for catching the misquote.

thereā€™s another reading. when i said LMD could have got you there i meant ā€œthereā€ = to the point you made in #21 where you describe historicism as having multiple themes rather than one clean criterion. LMD asked three questions probing those same distinctions. no reply, no further discussion. that conversation could have reached the same nuance months earlier.

thatā€™s a conversation getting dropped not you resisting error correction.

is this reading at least as plausible as yours?

54

i retracted acknowledged in #48 and clarified the point as an interpretation of #21, not a claim about what you said. that was a wording error i corrected in the same thread; #49 still centers the original wording rather than the clarification.

to expand on #53: this attributes a stronger claim than i made. the begging-the-question charge only lands if i was treating LMDā€™s unanswered questions as a completed discussion that established you were wrong. i wasnā€™t. the LMD point stands independently of whether you acknowledged anything in #21; itā€™s about three questions that didnā€™t get answered. there was no back-and-forth and no formal debate request; thatā€™s part of my point. LMDā€™s questions probed the same multi-theme / not-cleanly-historicist nuance you spelled out months later in #21. a dropped exchange, not a completed verdict.

i wasnā€™t presenting a concluded verdict from a finished debate. i was giving early diagnostic feedback about a possible forum failure mode. requiring feedback like that to come with prior concluded discussions, a formal debate request, or a worked-out alternative is a strong filter. youā€™ve worked on forum-design and debate problems publicly for over a decade; that is a high bar for early feedback from a new participant.

also, formal debate doesnā€™t seem to be the normal participation path here. the debate category has had no activity since aug 2024. so treating early feedback as a debate-request issue sends it to a path people arenā€™t using.

the specifics were not a finished proof, but they were there: LMDā€™s three unanswered historicism questions, the capitalism-thread interaction where i said your reply felt closed, and the specialist creators framing about people not posting because they risk criticism while you outclass people at debate.

your recantation thread raises the same category of problem. in #6 you wrote:

and:

thatā€™s the category iā€™m trying to discuss. #49 evaluates my feedback by standards suited to more formalized claims. if feedback like that is described as a lot of negativity and answered by pointing to the absence of prior conclusions or debate framing, then the forum may filter out the exact kind of criticism you later said was missing. that is the filter dynamic iā€™m trying to point at.

iā€™m not saying short replies or non-replies are inherently illegitimate; they can be reasonable bandwidth choices. the point is the secondary effect: if informal feedback is later treated as negativity or as improperly raised, the cost of half-formed criticism goes up.

and yes, these are several points, but they are all connected by one issue: whether this forum handles early, informal criticism well enough.

when i saw your microblogging post about tagā€™s reply, my takeaway was that you liked replies that engage specific quotes and explain the personā€™s thinking. #45 was trying to do that around one forum-dynamic issue, with examples. so seeing it treated as too many complaints to address simultaneously is a conflicting signal for me as a new participant.

you point to the same issue in your specialist creators article:

One of the reasons people donā€™t post much at my forums is they donā€™t know what they can say without a risk of receiving criticism.

and:

I also started outclassing people at debate too much and they donā€™t actually value losing debates in clear, conclusive ways (thatā€™s something I value highly but have nowhere to get).

that is close to the dynamic i was trying to point at. not a claim that you are irrational, or that a debate established you were wrong. more like: maybe some people do care, but the interaction cost is high enough that they stop trying.

also in #35 of this thread, when ActiveMind asked how to think about criticism usually being thought of as negative behavior, you said:

if respectful criticism about forum dynamics is reclassified as negativity or routed into objections about prior conclusions and debate framing, the high-status path you described in post 35 isnā€™t actually open for this category of feedback.

that list mixes actual agreement, tentative buy-in, questions, and ordinary conversational acknowledgments from a learning thread. none of it makes a later concern about a forum dynamic inconsistent. the like on post 221 was a local thumbs-up: it acknowledged your bandwidth/reply-style point, not the absence of any broader forum-dynamic issue. that same post 221 also said:

thatā€™s what i was trying to do. i brought up a related issue again with more analysis and a different angle.

i do have other communities with people much further along in their fields where informal discussion has been productive enough to change what i think is possible for me to pursue seriously. for example, one is quant-focused and helped me move from background reading to building much faster than i expected. they arenā€™t philosophy forums, so they donā€™t settle the comparison you asked for. but they are relevant to the narrower issue here: whether a forumā€™s interaction style changes how quickly people can learn from each other. also, even if this is the best available philosophy forum, that doesnā€™t answer whether this interaction has this specific failure mode. the question isnā€™t the total weight of the alternatives but whether this interaction style has a decisive error for the goal of receiving early criticism.

this exchange is an example of the cost iā€™m describing. #45 was a constructive observation about a possible forum failure mode. answering #49 has required a long procedural defense before the forum-dynamic point can even be evaluated. thatā€™s energy-consuming, and a formal debate would make the cost higher, not lower. if thatā€™s the path required for informal feedback to count, itā€™s understandable why people donā€™t attempt it.

55

The examples I quoted werenā€™t respectful or constructive. Letā€™s look at one of them:

Previously:

Now:

You basically called me below average, at some philosophy skills, compared to the general populationā€™s average. You explicitly werenā€™t just comparing to an above average group like your social circle, Reddit users or intellectuals. That doesnā€™t make sense as early diagnostic feedback because itā€™s extremely negative. You should be more charitable when you first see a potential problem. Itā€™s the kind of conclusion you get after lowering your opinion several times, but that conflicts with the early feedback claim.

Itā€™s also unhelpful as feedback given the lack of quotes, examples, specifics, reasons, suggestions, etc. Itā€™s not actionable or refutable. Early feedback is normally posted directly in response to one of the first few problems. The lack of specifics is more typical of late feedback where one is going by memory about multiple past issues.

Another way itā€™s late: you had already posted ~90 times and said positive things that led me to believe you considered our discussions productive.

You also appeared to save up multiple grievances and share them simultaneously, which is not how early feedback works.

What do you mean by ā€œmuch further along in their fieldsā€?

56

the full line began by framing this as an honest observation from my experience so far. i meant a local interaction point: the back-and-forth here has felt higher-friction to me as a new participant. i wasnā€™t saying youā€™re below average at philosophy. productive back-and-forth is about the interaction pattern, not a ranking of philosophy ability.

people with much more domain experience than me.

i donā€™t think saving up grievances is the right frame. it took a few interactions for a pattern to become visible, and i was trying to name it. this is also part of the cost i was pointing at: instead of the forum-dynamic point, we are now spending this reply on whether one sentence was too negative, too late, or not specific enough. i donā€™t think that answers the forum-dynamic point.

57

Do you utilize AI in your posts here?

This post, and several others, strongly seem to match the cadence of AI writing. I do not consider myself an expert in AI, so Iā€™m curious if others have noticed this or not.

Iā€™ve also noticed that your lack of capitalization seems highly consistent, which is a bit odd to me. You do not capitalize first person ā€œIā€ and you do not capitalize the start of sentences, but you do capitalize names and acronyms.

In my experience, lack of capitalization is more ad hoc due to posting when low energy/low effort. Seeing it follow such strict rules is odd and makes me think of an AI prompt that included instructions not to capitalize.

2 Likes
58

Yeah, I notice it too, itā€™s not just you. I think Iā€™d be willing to bet money that a good amount of it is AI generated. Thatā€™s how ā€œconfidentā€ I feel about it @Oracle.

Yeah, I think it was on this forum that another member posted about a Jack Dorsey tweet being AI, and Jack didnā€™t capitalize either. One of the comments under Jack Dorseyā€™s tweet called it out I believe. Something about it usually taking more effort not to capitalize than to capitalize. I then tested it and when I prompted the LLM (I think it was ChatGPT) to write a response more casual or something, it wrote it without capitalizing, in the same style as Jackā€™s tweet.

To be fair though, AI use is becoming more and more common. My guess is everyone on the forum uses AI to some extent in their life (maybe not to write but for other common use cases) I think itā€™d be valuable for @Elliot to write an explanation about when AI is good to use for writing, thinking and philosophy, and when it isnā€™t and why. Otherwise it could be misinterpreted as trying to police someoneā€™s ways of learning and expressing ideas.

Before I joined this forum as a poster, I was a long term reader of it. When I made my first ever post on here, I was so concerned about being grammatically correct (I was under the impression @Elliot was strict on that for some reason back then) that I copy- pasted my rough draft into an llm and had it proof read and edit my words, then pasted the corrected/revised version the llm generate directly to the forum. I remember not liking the style it wrote in though.

Even telling you folks that makes me feel a bit weird though. Part of me feels like Iā€™m confessing a sin or something lol, and I know that effect is not anyoneā€™s intention. Thatā€™s why I think a more detailed explanation on AI use on the forum would be valuable. If youā€™ve already written one @Elliot, I havenā€™t seen it.

1 Like
59

@anonymous105 Iā€™d value your explanation on good and bad ways to use AI for writing and philosophy as well.

60

What could be misinterpreted? This forumā€™s rule is to disclose AI writing in forum posts, not to prohibit AI use.

FYI I think having an AI rewrite anything (or otherwise copy/pasting AI text) should be disclosed. Whereas if you ask an AI for typos and grammar errors, and then you fix them yourself, you donā€™t need to disclose that.

Yeah that essay could be interesting. Iā€™d be interested in other peopleā€™s thoughts on the topic too.

1 Like