JustinCEO Topic

This is a topic where JustinCEO can post stuff while other stuff is unresolved. We’ll see how this goes.

Other people can talk here too. And JustinCEO may to reply to stuff from elsewhere by posting here.

1 Like

Thanks for the topic @Elliot :+1:. My current plan is to try to comment on the Pinker video when I have some time and energy. I’ve had thoughts about other threads but I think I’ll mostly keep them to myself for now. Overall I am trying to do fewer things more consistently.

In another thread @Lebowski said

  • Regardless of format, text analysis is the CF topic I seem least interested in. I like other topics you write about more.

I’m curious why that is. I enjoy text (and grammar) analysis in general, and have enjoyed participating in it myself in the past. I didn’t comment on the video being discussed in that thread, but that doesn’t mean I didn’t enjoy it. My lack-of-commenting issues are a separate issue from whether or not I enjoy something. I actually think I might be more likely to comment on something if I dislike it, at least in certain circumstances. Haven’t thought that through.

I don’t know all the reasons but have some guesses:

  • Explicit text analysis & grammar remind me of grade school, which I mostly had a bad time in.
  • The application of literal text analysis carries negative value in many of the areas of my life.
  • I think I’m good enough / reached the point of diminishing returns in the areas of my life other than FI where literal text analysis carries positive value.

Can you concretize this one a bit more? I’m not sure what you have in mind.

My guess: by “carries negative value in many of the areas of my life”, do you mean that if you applied literal text analysis to say, work emails or texts with family, it’d cause big problems for you?

Ya, although “big” is overstated. It causes friction, makes relationships harder, wastes some time, that sort of thing. It doesn’t blow up my career or my family life (which really would be “big”).

As context: I think I have or have had thoughts along the same lines as you, which is why I’m interested in the topic. It’s also easier for me to analyze and potentially criticize someone else’s thoughts on the topic since they will likely have somewhat different thoughts (and I’m biased about analyzing my own thoughts.)

To clarify: do you think having the knowledge of how to do literal text analysis would inherently cause you problems in these cases you are worried about? Cuz e.g. it would just start happening automatically and you would judge people poorly as a result and have problems from that? Or are you thinking of trying to engage in literal text analysis together with someone and that causing problems (cuz e.g. they’re incompetent to do it and get angry)?

Also to clarify: I asked my question in the form of a hypothetical, but you wrote your response in a manner that suggested that perhaps you have already tried applying literal text analysis in certain situations you describe and you noted these problems. is that accurate?

Discussing problems with shipping companies, @internetrules said in another thread:

One thing you could do is never rely on shipping companies to either pack things or even to correctly put labels on them.

When I ship stuff, I print out the label myself (on the carrier website, at a third party site like https://www.pirateship.com/, or from the store website if it’s a return), affix it to the box or envelope myself, and just hand the carrier the finished, sealed/taped, labeled shipment. You can do this even for stuff like next day service (though for certain services you may need to use a special envelope or box). Yes, they could still smash it or misdeliver it, but I’ve at least eliminated one potential set of things that could go wrong. It’s very sad that you can’t reliably outsource putting stuff in a box or envelope and putting a label on it.


Analyzing what someone said literally is a bit like a spoiler to a story: You can’t unsee it once you’ve seen it. It’s theoretically possible to ignore the analysis and focus solely on guessing what they consciously expected you to understand. But that’s hard in practice.

It’s also possible to know how to analyze literally but then choose not to do that in a particular situation. That’s easier even though it does still take some effort if someone’s conscious expectation isn’t easy to guess.

Not what I had in mind but I think that’d go badly too.

Ya. I’ve had people get mad at me for analyzing their words literally.

What people in some situations actually want is for you to know the social script they’re playing, and use what they say as a time sync to that script so you can play your role in it correctly. Analyzing what they literally say tends to take you off script, which they don’t like.

This reminds me of a problem I’ve had related to being too smart or too good at analysis.

I’m very good at tests in general, but sometimes I struggle with ambiguous questions, while many other people don’t notice they’re ambiguous and just read them one way. Their one reading is usually the same way the test designer meant it, especially for the people who are pretty good at tests.

If I see that a question wording is ambiguous, I can’t unsee that. I have to e.g. mentally model the test creator and figure out what they would have meant or intended, or mentally model a typical person and figure out how they’d read it.

This is related to recent events at EA: I quickly noticed quoting as a method of posting text on the forum that isn’t CC BY licensed. I tried to get clarification about whether it works how I think it does and also whether it’s OK to use much or they’d prefer we don’t. They refused to clarify. I guess noticing that edge case was too clever for them and they just found it alienating and annoying. I thought it was a good, useful, helpful, important point that could contribute to making the rules clearer. I didn’t realize at first that it was too advanced for them so they’d react negatively instead of being happy to clarify their intention or preference.

I agree with the part about being unable to unsee it.

Here’s what I think you’re saying: the default, conventional way of social interaction works okay for avoiding certain bad outcomes. If you learn how to do text analysis, that adds more information to deal with, and that layer causes problems by preventing you from following the normal conventional way of social interaction easily. The information also doesn’t meaningfully help with doing social interaction. You can willfully blind yourself to this information, but it’s pretty hard to do that. So it’s best to not have the information from the outset, to not develop the text analysis skill, and to thus be unaware of the literal meaning of what other people are saying to you. Is that a summary you agree with?

I’m unclear on your point here regarding the relationship between choosing not to engage in literal text analysis and someone’s conscious expectation being hard to guess.

Assuming I’ve gotten the gist overall so far, I have the following tentative disagreements:

  1. I agree that there’s an issue regarding dealing with the new information one gets from literal text analysis, but that seems like a skill issue that can be managed with practice. You use language like “theoretically possible” which makes me think you think the issue is inherently super hard. Feel free to clarify here because maybe we don’t disagree as much as I think we do.

  2. People following conventional social scripts and interaction rules have things blow up pretty often, so it’s not super safe to just do that. You seem to be assuming that’s a safe fall-back approach.

  3. I think that having the text analysis information is ultimately good, and tells you more about what the world is like and the meaning of the sorts of things people say. If somebody in your life routinely talks nonsense (from a literal analysis perspective), that’s important to know. It should affect not just your opinion of them but also the kinds of projects you engage in them with. So it’s practical, useful knowledge to have. You seem to think it’s not helpful knowledge.

What do you think?

I’m still planning on analyzing the Pinker video but have had greater-than-expected-work-related-unavailability so I’ve been delayed in doing so.

Note for all my Pinker video comments that I’ve seen the video before so I have prior exposure to the content.

This is the tree for the first sentence that Elliot makes in the Pinker video

This is the one I made after pausing the video and before looking at Elliot’s answer:

The only notable difference I see is where I attached the “that” (and that is something I debated, so that’s pretty expected).

In the discussion Curiosity – Controversial Activism Is Problematic, @Lebowski said:

I don’t know how Elliot would address this but I think I have a thought worth considering.

Consider some project where there’s absolutely no opposition by anybody that affects the effectiveness of your efforts. For example, consider the project of spending $100 on donating some philosophy books to kids who requested the book (this is, or has at least been, a real thing). Nobody’s fighting you on that directly, so you can just spend whatever without obstruction and maximize the value you get. Your dollars are fully 100% effective at buying you the books at the market price to give to people.

Now suppose you are considering donating to a controversial political cause, like say something related to guns or abortion. The efforts of the other side of that debate will reduce the effectiveness of your dollars. Not only that, but various factors completely outside of your control (like events that are interpreted as reasons to pass new legislation, general political dynamics and cultural trends in the country, the ideologies taught to people at school) will have a massive impact on the effectiveness of your donation as well. Your $100 is a bit like a single vote or maybe a few votes in a Presidential election in terms of its importance.

Suppose you are considering donating to the book cause but then donate to the political cause instead. What is the effect on the book purchases? Books are pretty cheap, and nobody’s fighting you on this, so your money probably makes some difference here. Some book purchases probably just won’t happen.

Suppose you are considering the political donation but then switch to the book purchases. What is the effect on your favored political cause? Probably nothing. If $100 was enough to notice at all, it’d have to be a pretty small group that probably wouldn’t have much influence anyways in a controversial political fight. If there is a general downward trend in fundraising, well-organized groups have ways of dealing with that and stepping up their fundraising anyways.

So I guess my point is that you have to consider the marginal impact of your $100 in the different causes. The marginal impact is going to be highly variable depending on stuff like whether it’s a controversial political debate with two generally pretty well-funded sides and other factors.

Suppose you don’t donate $100 to the controversial political group, and the other side in the debate wins a big political fight and passes some bad laws. Should that whole harm be ascribed to your failure to donate? No. That’d be kind of like you blaming every bad thing from some Presidential administration for your failure to vote against him in his elections. But where you said…

So suppose I choose A because it’s uncontroversial and my efforts there won’t get cancelled out. Life gets 100 units better because of my work on A. But because I left B unaddressed, life gets 100 units worse because of anti-B’s work. A and B cancel each other out. I’m still left in the situation where total life benefit units don’t change.

…you seemed to be doing something like that. I think you were assuming that you could just move resources around from non-controversial to controversial efforts and have them be the same level of effectiveness. But I don’t think that actually makes sense, for the reasons I’ve explained above.

Given the same resources or efforts, the marginal effectiveness of expending them in a controversial fight is going to be reduced compared to spending them on something else non-controversial. It’s sort of like a reverse multiplier effect of effectiveness (reverse if you think of it as going from non-controversial to controversial; it could be just a regular multiplier effect if you go from controversial to non-controversial).

By the way, I think the same point applies if you’re considering donating to a controversial political cause versus investing in yourself (in terms of professional education or personal development or whatever). Investing in yourself tends to not have an active opposition (kinda funny to imagine that existing), so you don’t get this reverse multiplier effect thing if you do it.

Note that the analysis is very different if you’re actually in a unique position to influence the outcome of, say, some controversial political fights. If you’re a sitting President, Senator, or Governor, then while it’d still be ideal to reach broad agreement where you can, controversial fights are inevitable and you matter to them. In a sense, you have unique or super high value specialized political resources: formal political authority, the ability to get news attention, and so on. That’s not the situation of the overwhelming majority of people though.

More Pinker video. My diagram of Sentence 2 was identical to Elliot’s.

My diagram of Sentence 3 was almost identical, except I represented “that” twice as both a relator and noun. I think we basically agree though.



More Pinker, sentence 4:




  • Elliot nested the “by physical operations” phrase under “transformations” instead of under “knowledge”. I went back and forth on that a few times myself.
  • I represented the “that” twice again, and Elliot didn’t.
  • Elliot grouped “are designed to preserve” into one node while I broke it up.

None of these seems fundamental.

More Pinker, Sentence 5



  • I attached “to effect” to “operations” instead of “control”. (I went back and forth on where to attach it).
  • I wrote out what I believed to be an implied “that are” before “guided.” (I’m not sure about that one).

More Pinker, Sentence 6




  • I treated “that achieve intelligence” as a clause with a subject, verb, complement.
  • I put “the” directly as a modifier on systems.

I think that what Elliot wrote is actually compatible with this approach when you take into account his color coding key, but my intuition wanted to represent it more directly in the structure of the tree itself.


More Pinker stuff tomorrow hopefully.

Just thought this was interesting.
CVS committed to donate $10 million to the American Diabetes Association as part of an agreement, then asked customers for donations to the ADA at checkout. The claim is that people reasonably thought a donation would result in an extra dollar going to the ADA, but really it was just reducing the amount out of pocket that CVS had to pay anyways. CVS is being sued for fraud and that seems reasonable to me.

One criticism about the twitter thread: for some reason the guy brings up CVS’ revenue to criticize them and say they should have just paid the amount, but profit is really the relevant figure IMHO. It would still probably be a plenty big number, though of course not as big.