Elliot's Microblogging

So far (3/4 through round 2) I don’t like these broadcasts as much as the St. Louis Chess broadcasts because they go into less depth about individual games. It’s a larger tournament but I’d prefer if they’d focus on the top boards or best games more instead of trying to cover every game some. I’m hoping they may focus more in later rounds when there are clear contenders to win the tournament, and therefore some key games.

Also, I like the St. Louis segments where the two main broadcasters go to Maurice Ashley who has been using a strong chess engine to find the right moves while off camera, and he goes over some of the right answers in interesting positions. The chess.com broadcast is making little use of computer analysis to reveal correct answers to some of the hard positions.

Also, St. Louis has better roles. Yasser Seirawan (despite being older, wiser, highly experienced) plays the role of the more naive person who wants to try things out. He asks some of the questions viewers would have and is the most relatable for viewers. Then Alejandro Ramirez or Ashley can act as his foil and give criticism of some of his experiments.

On the chess.com broadcasts, Anna Muzychuk is doing both roles: she’s the primary person analyzing good lines and giving answers about what works or not, and she’s the primary person experimenting with moves that might be bad, trying things out, asking questions, etc. She’s the more active person who talks the most and also does the most chess moves and has the most chess knowledge, while Stuart Conquest isn’t contributing as much.

It’s pretty standard on two-person broadcasts to have one person who is more social and one who is more of the expert. People use these roles on purpose even when they could play a different roll. E.g. I remember a different chess broadcast where a woman kept doing the social stuff but then when a guy left temporarily she immediately switched into acting more like a chess expert until he got back. So doing the more social side was an intentional choice/role, probably because she was better at it than the men, not because of lack of chess ability by her (I do think the men were stronger chess players, but she was plenty strong – far far stronger than the audience including me, and capable of playing an effective chess expert role).

Similarly, on Starcarft broadcasts, Tasteless intentionally plays dumber than he is and acts like more of a viewer representative and social/fun guy, while Artosis plays the expert role. The roles are intentional and help organize who says what kinds of things. Tasteless helps support Artosis’ expert reputation and tries to avoid challenging or undermining it. Tasteless will sometimes ask questions for Artosis to answer, even though he knows the answer and could answer it himself. Artosis is somewhat more of a smart expert than Tasteless, but they play it up and have some division of labor. (Possibly, over the years, they actually became more different due to playing different roles on broadcasts. Tasteless had less incentive to keep studying expert stuff, and more incentive to learn how to do his role well, e.g. getting better at making jokes, filling dead air time, or understanding and remembering to speak to the perspective of audience members who don’t know a lot about the game.)

I currently have 5 articles scheduled to post on the CF site. They will continue to be on Thursday and Sunday mornings for now.

1 Like

There’s something bad and unrealistic about this article. Why does Reisman think Fox is or should be good? They’re just shitty. Any network could improve. Why focus on Fox improving? Why not ignore Fox and seek progress elsewhere? I think Reisman overestimates Fox.

lol I was reviewing old exploratory writing to find something to edit and I found an 865 word paragraph.

That’s so long it’s funny, but then I read read the beginning to see what it was about:

structuring [knowledge] as many small parts with connections

I laughed at the irony.

(I knew the paragraph was way too long when I first wrote it. I just figured I could split it up during editing later. Sometimes I have ideas flowing and just write them down quickly without getting to an automatized paragraph break. Usually I separate out paragraphs while writing automatically but the automatic rules don’t cover every case. So it’d take some attention away from writing my ideas down to split the paragraph using conscious thought about how to handle it. In that case, it’s often better to do that conscious attention later instead of letting it distract from the topic I’m writing about.)

Sad seeing one of the world’s best players trying to play super fast chess games using software that’s worse than we had 20+ years ago.

He doesn’t have smartmove. Normally to specify a chess move you have to input two squares (where you’re moving from and to). This can be done with click and drag (as in the video) or two clicks.

Smartmove is a feature where you click on a single square and if you have exactly one legal move involving that square, the move is played based on just one click. So if you click one of your pieces and it can only move one place, it moves there. If you click an empty square or square with an enemy piece, and you have exactly one piece which can go there, then you play that move in one click.

If you do a single click on a square that works for 0 or 2+ legal moves then smartmove has no effect.

Smartmove is a significant convenience and speed increase, and he’s getting into many games where moving just slightly faster would make a significant difference that would help him win (or if his opponent had it too, it’d help both of them put on a better performance – play more moves faster. it’d help their hands keep up with their minds better.)

I think the problem is the newer chess websites force you to play with their shitty software in a web browser. And I think none of them support the wild 5 variant or various other chess variants that we also had 20+ years ago. Yet these newer chess websites are now dominant in terms of popularity and having strong players use them.

Old chess servers let you connect with telnet to play with an ASCII board, and anyone could write software with a graphical chess board which connects and plays moves over the simple protocol. The servers were designed with an open protocol to be compatible with whatever chess playing software anyone wanted to create, rather than forcing you to use a single piece of software.

I recall there was a chess interfaces named “Thief” because it purposefully copied/stole the best features from all the popular chess interfaces at the time it was made. This was, of course, a good thing. It was made by a strong player who played fast games. Both smartmove and premove were innovations that people came up with in I think the late 90’s and which ended up in multiple apps because they were useful. Thief had both though I don’t think it originated either. Premove is still around but somehow smartmove got forgotten or something. (Premove is making a move on your opponent’s turn. If it’s a legal move when it’s your turn, you play it instantly. In most positions, premoving has a risk of playing a very bad move because you can’t react to a threat your opponent just made. Premoving recaptures is routine because they aren’t a legal move unless your opponent plays the capture. Premoving is also common when you have only a few seconds left and there aren’t many pieces left. A common alternative to premove with lower risk is, on your opponent’s turn, dragging a piece to a square and keeping your mouse held down. Then as soon as you see his move you let go to move, unless he did something unexpected that makes your planned move really bad.)

EDIT: Here is someone 2 years ago asking for smartmove as a feature request at one of the current, popular chess websites: Smart Move • page 1/1 • Lichess Feedback • lichess.org (he got some dumb replies and seems to have been ignored).

1 Like

I don’t know about chess software specifically. But something I’ve noticed in general is tons of software and service development focus and effort has shifted away from empowering the user to do what they want and then (maybe) charging the user for that value. Instead, tons of effort now seems to go into nudging, dark-patterning, de-featuring, and outright coercing users into doing what you (the developer or, more often, the developer’s big-tech employer) want them to do so as to generate as much revenue as possible from some combination of direct user payments, advertisers, content owners, and venture capitalists.

A couple of areas I’ve especially noticed it in -
Music - 20 or so years ago Microsoft (and others, but I’m most familiar with Microsoft) were putting lots of development effort into library based music players. For Windows, this was Windows Media Player. The user owned the music (as files in an industry standard format). The software gradually got better and better at letting you do whatever you wanted with that music. The best feature was automated playlists with programmatic rules. When combined with individual, granular (star) ratings and tags for genre, sub-genre, mood, etc. it became possible to program an automated DJ that did exactly what you wanted, nothing more/nothing less. If it played a song you didn’t want to hear at that time, you didn’t just have to hit “skip” and hope some opaque and other-controlled algorithm insured it didn’t happen again. You could figure out why, and update the rules of the automated playlist so it didn’t happen again. If you heard a new song you liked on the radio, you could buy that song, tag it appropriately, and then it’d show up in all the correct automated playlists. You could keep running an old version of the software as long as you saw fit, and you could change player software itself (or even write your own) without re-buying all the media files, knowing all your music would play in the new software.

10-15 years ago Microsoft quit adding features to Windows Media Player. Nevertheless, it’s still what I almost exclusively use for playing music. I’ve expected any day to hear they are dropping it. Instead I recently heard they’re going to have a “new” version soon in Windows 11. I don’t have high hopes but we’ll see.

What seems to be getting most of the attention now is streaming services where as a user you don’t own the music and your control over what plays is limited. Sure, you can play an individual song or create a static playlist and you can thumb up/down songs. But (at least from what I’ve seen) you can’t exert the kind of absolute control over your music experience that a good library player like Windows Media has. Thumb up/down in particular is pretty useless for anything other than “I never want to hear this song again”. And you have to pay every month or you lose access to it all. And for everything non-static it gives the service an “in” to play you…whatever the service thinks will maximize its revenue. Which may or may not be what you actually want to hear at that time. And if you want to move to a different streaming service good luck taking your ratings (such as they are) with you - you have to start over trying to get enough metadata into that system to get it to behave the way you want. Also, maybe not all the music you like is there and if not it’s not practical to add it into your mixes. And the services can & do change their interfaces and features - as a user you have no control over if or when this happens.

The only advantages I see for users in the streaming services are not having to buy a bunch of music you like up front and not having to store & manage music files yourself. And I’m aware neither of those were issues for me by the time the streaming services came along but maybe they are significant issues for lots of people. So it’s entirely possible the streaming services work better than library players for the majority of people. But the lack of ownership and control in such services is also glaringly obvious and I don’t think it’s accidental.

Video - Netflix is the main example here although my impression of other video services is broadly similar. As a user you never owned videos at Netflix - something I found fine as (unlike music) I only rarely want to watch something multiple times, years apart.

When Netflix started it had two very cool user control features: A queue, which you could order as you wanted. And (like the music example) star ratings of what you’d previously watched (whether on Netflix or elsewhere). Once you’d rated enough videos, Netflix would get pretty good at estimating the star rating you’d give to other Netflix content you hadn’t yet rated. The algorithm was opaque, but at least it worked in so far as the predictions seemed reasonably accurate. You could then add a video to your queue (or not) and even put it in priority position based on the predicted rating. So when you had time to watch videos, you’d go to your queue and your highest priority items (Meaning: stuff you’re most likely to enjoy) would be right at the top.

Some years ago they did away with the star ratings in favor of (like the music streaming services) thumb up / down. Way less informational to their algorithm, and at the same time they dropped the predicted rating that made their algorithm particularly useful. Instead, they suggest shows based on general criteria like genres & popularity, along with some (opaque and in my experience not accurate) effect from what you’ve thumbed up or down.

But at least they kept the queue, and I could guess (not as well as the old algorithm, but better than random) how much I’d like a video and put it at the right place in the queue, and only watch stuff from my queue as opposed to what the service prompted on any given day. Lately they took even that feature away, removing the ability of a user to order their queue as they see fit in favor of an algorithmic queue ordering. So now what’s at the top is explicitly not what you as a Netflix user think you’re most likely to enjoy watching. What’s at the top is whatever the algorithm wants to nudge you to watch.

I’m not sure if, taken as a whole, the no-rating / no queue-ordering is better for Netflix’s subscriber retention than the rating system and self-ordered queue. I think it’s possible - maybe people want to be more passive about what they watch than I do. Or maybe not.

But I’m quite sure it gives their execs much more power over what people actually watch than the old system. It feels super manipulative, and my guess is that’s why they did it. I’d be surprised if it’s not heavily influenced by stuff like what agreements they have with content providers, what shows they’re trying to create “buzz” around, and other factors that I as a user do not want influencing which videos I watch. I have to keep my own list outside of Netflix to maintain the level of control over my video watching that I had under the old system, and near-zero people do that.

In both cases (music and video) I think it’d be easily possible to keep user control features intact while providing an “easier” algorithm-guided interface to users who didn’t want the control. But the developers (or their employers) seem to actively not want that. It seems like they’d explicitly rather manipulate users than give them control.

I don’t know enough to say whether this user-manipulation is economically productive or not. I just know as a user I notice it, find it super annoying, see it as a general industry trend and I’m skeptical about it being good in an economic sense.


Search results on a lot of websites are quite bad. I have often searched on Amazon for a book that I know exists with the title and author and it’s not among the search results. So I have to go to a non-google search engine to find the book on amazon.

Attention seeking is problematic in a deep way b/c error correction is attention. So errors (at all other goals) suffice for their meta goal.

Something that works particularly well, for getting attention, is errors at goals other people care about.

I’ve uploaded importable RSS feeds for my stuff. The picture shows what’s included:


Elliot RSS feeds.opml (1.9 KB)

Another example is smart TVs. You used to have a plain old generic TV and hook up accessories like a VCR, dvd player, tivo, roku, apple tv, etc. You also could hook up a computer to the TV or a video game console. Now they build in a specific type of smart TV accessory, and opinionated take on software, right into the TV itself. You can still hook some things up but they are moving away from that and trying to control what you watch and how you watch it. They are trying to move away from the model where you choose your own accessories.

They’re working on streaming gaming services now too, kinda like netflix with games instead of movies/tv. Apparently having Steam control all your games wasn’t a bad enough transition and they don’t want you to have a local copy of the game on your computer at all. Instead they’ll run the game on their computer and you can just send your inputs over the internet and they’ll send you back video and sound from what happens.


Old meme. I think Tesla and others (quite possibly Apple) are trying to make some of it a reality. Besides cars, companies have been doing similar things to TVs and various appliances. I actually just threw out a kitchen appliance partly because the maker went out of business and stopped updating the iOS app. They’re integrating bluetooth and phone apps into toothbrushing now. For my sous vide stick, I purposely got one with physical buttons with merely an optional phone app, but I don’t know if those will still be available in 20 years. The Joule, which I didn’t get, is smaller. It does have some advantages due to skipping the buttons and skipping the small screen that displays the current temperature setting.

I have a particularly negative opinion of the hardware that requires an ongoing paid subscription for the software to keep working well. E.g. peloton exercise bike. I think the physical objects I stick in my home ought to work fine with no active subscription (and when my internet connection is down) unless there is a very good reason for an exception (e.g. my cable modem won’t work effectively without a cable internet subscription). Subscriptions can be for extra features like workout classes and videos, but IMO your actual gym equipment ought to work by itself.

Maybe in a decade we’ll get some really dystopian stuff like home gym equipment that requires an active internet connection to use so that they can track how much you exercise and set your government-mandated health “insurance” rates accordingly. Or maybe you can use it offline but it just don’t count towards your insurance rates. People who go exercise outside will need to bring a smartphone and smartwatch in order to get any credit for their exercise. Obviously some types of exercise will count more/better than others according to the mysterious software and there will be a whole industry developed around how to most efficiently lower your health insurance rates – which exercises count the most (regardless of how they help your actual health). The health “insurance” rates will be more of a tax than an insurance system.

1 Like

Geller Report has apparently switched its focus from Islam/anti-semitism/Israel type politics to COVID. They sent out an emergency email about how Facebook is going to deplatform them over COVID info (they have 1.5 million FB followers).

Why did they pick the COVID battle and care enough to switch topics? I thought they cared about that other stuff, prioritized it, and had built up an audience with the other stuff. But now they are fighting with Facebook over COVID…

What used to be a similar blog, Jihad Watch, has not switched topics. Both are associated with David Horowitz which is a sign he does not tell them what to write about (since one switched topics and one didn’t), so that’s good.

Genshin Impact rotates special bonuses and enemy types in the game’s hardest dungeon. they purposefully – and very obviously – make it favorable to the character they are currently selling

ppl then get the new char, go do dungeon, do well, tell ppl the char is good
other ppl say: hey watch out for this bias

but ppl are so bad at overcoming bias… so it works a lot

example: https://www.reddit.com/r/GenshinImpactTips/comments/riohc8/those_who_got_itto_did_he_live_up_to_the_hype/

it reminds me of ppl who “know” the NYT is biased but are still tricked by it. they partially correct for the bias but their correction is small/incomplete. (there are also ppl who hate the NYT so much and won’t listen to them at all, so they are overcorrecting. but it’s rare that someone actually can listen to some stuff while not being fooled much).

I deleted the politics section of the FI book recommendations page. I wanted to say “don’t read politics, but if you’re going to ignore me (as I know some people will), here are some tips”. But I think it’s better to refuse to say anything. People are interested in politics for bad reasons. Hating the outgroup makes your life worse. It’s also bad to be fooled into feeling like you have more influence or power than you do. Work on self-improvement instead.

1 Like

There are so many amazing perspectives [in these podcasts] that it’d take me a month to excerpt and comment on each.

Instead, maybe let’s do this…

If you hear something that doesn’t make sense to you, or sends you into a rage, or triggers a massive a-ha moment, please reply to this message with your feelings and I’ll comment and share with the list.

I think that’s an example of pretty good reply bait. I prefer not to write stuff like that though. I generally don’t like to spend time on it, and I also prefer to deal with people more capable of saying things on their own initiative.

I read over 100k words yesterday. (That would be around 350 paper pages.) Fun but tiring. I haven’t read that much for a while.

Also, unrelated, I’ve been doing a lot of editing recently. Many of my recent CF articles were written in the last 4 years.

Months ago, patio11 tweeted a bunch about his Vaccinate California startup/project.

The general impression I got was that anyone could have done it, just like with most startups. You just need to take some initiative and do something useful.

He did it because he was motivated, driven, put in the work, and no one else did it first.

The community flocked to help with donations of money and labor (from programmers as well as simpler stuff like people making phone calls).

In general, startup success has more to do with social networking than people talk about or admit.

But at least with a regular startup you’re selling something, and customers might like it and buy it regardless of who your friends are. And investors like to make money.

With a charity it’s much harder. There’s no big pay day. The users don’t pay you. You have donors instead of investors. Capitalism doesn’t help level the playing field.

patio11 had success where many other people would not have been able to do the same thing because of his social status and his contacts in the tech industry. It’s important that people observing be aware of how achievable a similar thing is or is not for them. patio11 is systematically misleading about that.

I’m fascinated by the idea that outside events can make you think something, literally against your will.

For example:

If I said to you, “Don’t think of an elephant,” you’d be forced to think of an elephant in order to comprehend the statement.

By the time you tried not to think of the elephant, it’d be too late.

And oh by the way, I just made you think of an elephant.

And maybe this elephant idea triggered a pleasant childhood memory of going to the zoo with your grandparents…

And then maybe that memory made you think of spending Christmas eve at their old house, and so on…

Why are people so confident that this trick always works? That it makes people have thoughts, even against their will?

It did not work on me when I read this text. I recognized the word “elephant” as a noun representing any noun, and I recognized what the sentence meant conceptually (I’ve seen it before), and I moved on. You can know what the sentence means without picturing an elephant.

They (people who do this trick – I’ve seen it before) don’t even say what they mean by “think of an elephant”. Do they mean picture one? Otherwise, how many traits of an elephant do you have to load into active memory for it to count? E.g. if you see the word “elephant” and remember or think of the trait “noun” does that count? What about the trait “animal”? Big? Heavy? Gray? Mammal? Ivory tusks? What about a combination of traits? What pieces of data/information count? But I suspect they just mean picturing an elephant, and this emphasis on thinking in pictures – which they take for granted – is related to how most people are kinda bad with words and logical thinking. A lot of people focus on visualizing things (and sometimes the other senses) and on emotions (they remember how stuff made them feel instead of what happened), and it’s hard to do logical analysis or deal with ordered lists of thousands of words and punctuation when that’s how you habitually think.

1 Like

Schools teach stuff that’s really contrary to Paths Forward.

You can’t hand in a published article to respond to an essay prompt. You have to write your own, even if you have nothing new to say and think someone else already said it better.

You can’t do it in a Paths Forward way either where you say that you take responsibility for the arguments in the article and you’ll responds to any questions/criticism, but after a literature review you were unable to identify any outstanding questions/criticisms that need answering. You could even turn in a whole tree with cites to the answers to 50 potential questions/criticisms, and you’d still just be given an F. Students are required to address topics in a non-Paths Forward way.

The emphasis on saying something yourself to get credit for it is a major cause/incentive for plagiarism too. Of course not every student can think of an original answer to every essay question. Most students responding to most prompts will be writing unoriginal stuff and trying to make it different enough not to count as plagiarism, but not too different – it needs to be similar to existing answers so your teacher can recognize it as a standard answer and give you a decent grade (even if you did have a important thing to say as an answer, that didn’t fit one of the expected answers, you might get a bad grade for it – in fact, with DD’s personal help, I tried that a few times with bad results from graders). And you aren’t supposed to quote too much or rely on cites too much for your main points – but at the same time you’re trying to say stuff that’s already been written while having no ideas about how to improve on existing publications by experts.

EDIT: lol a couple min after writing this I see this: https://www.tiktok.com/@bad_news_bares/video/7046451615649271046

1 Like

A post was split to a new topic: Reading Methods, Speed and Quantity