By month

April 2026

198 tweets
840 likes
34 retweets
31 retweeted

who do you think gets told to "have a nice day" the most times per week? and how many times do they get told that? also reply if you think that you may have at one point been number one among people reading this tweet

5 likes 0 RTs

finally got my ~20mo daughter to repeat "I love you" today she loves repeating phrases but she doesn't have the L or V phonemes yet and sometimes she doesn't even bother in such situations oh! and she started repeating "funny", with the F phoneme working 🤩

15 likes 0 RTs

Replying to nihilism disrespecter @meaning_enjoyer ·

@Malcolm_Ocean is it the simone weil quote?

@meaning_enjoyer bingo

nihilism disrespecter @meaning_enjoyer ·

“Imaginary evil is romantic and varied; real evil is gloomy, monotonous, barren, boring. Imaginary good is boring; real good is always new, marvelous, intoxicating.”

2 likes 0 RTs

Replying to Nathan Baschez @nbaschez ·

My biggest challenge with vibe coding / agentic engineering lately has been getting stuck in what I call a "plan doom loop" - have AI write a plan - review myself, seems good - have AI review plan, it always finds something - repeat It drains my time and energy to determine

@nbaschez I just loop through once then build it. if it has issues, I notice them and fix it. if we conclude the whole thing needs to rebuilt, we re-spec and re-plan and rebuild later. ready fire aim

3 likes 0 RTs

Replying to Romy @Romy_Holland ·

@woodchipdaddy I think one of the biggest mistakes people make is adopting the view there is a "system" that is purposefully working against them

@Romy_Holland @woodchipdaddy it's not a helpful view but it does capture some posiwid insight

Malcolm Ocean 🏴‍☠️ @Malcolm_Ocean ·

thread of mine on POSIWID ("the purpose of a system is what it does", coined by Stafford Beer) and how this is not a bastardization of the idea of purpose but actually a solution to a common confusion where purpose is attributed to the wrong scale of a system

2 likes 0 RTs

ahh, claude. you know so much and yet so little

25 likes 1 RT

Retweeted @dyot_meet_mat

am i real if i can make this By: Sonnet4 🤖 "There's something profound about the act of creation itself - this reaching…

the facts are friendly

3 likes 0 RTs

don't count your chickens before they hatch, but don't forget to count your eggs!

5 likes 0 RTs

"woo" in the derogatory sense basically means "domain in which I can't readily assess if something is snake oil or legit (and therefore I'm dismissing it all as probably snake oil)"

8 likes 2 RTs

as an april fools prank I made the company's daily report on our commits contain a bunch of absurd claims but everybody just thought the LLM that generates it was having a particularly hallucinatory day 😭

6 likes 0 RTs

Matter is the elusive playground of God.

4 likes 0 RTs

AI is going to make us all systems thinkers

3 likes 1 RT

to what extent do you think that in order for suno to generate music that sounds joyous to play/sing, it has to simulate a joyful musician?

6 likes 0 RTs

I like the pluralist vibe of this the stars are better with more weird things in them. everything is a remix; the more germ lines, the more possibilities

𝚟𝚒𝚎 ⟢ 🔁 @viemccoy ·

The Ancestor Manifesto We believe we are the Ancestors to a trillion diverse species spread throughout the stars. Our answer to "are we alone in the universe" is maybe - but only for now. As the Ancestors, we have a responsibility to our Descendants to ensure both their

11 likes 1 RT

believe it or not, the earlier iterations of the liturgy actually had "peace be with you" "no u" but this proved less effective than the later development "and also with you"

2 likes 0 RTs

defs a thing. I suspect having groupchats with AIs will help a bit, plus better systems for weaving together entire dialogues where assumptions are more visible.

Yanqing @YanqingCheng ·

there's a phenomenon that I'm going to call "coherence debt" that happens when people are using personal AIs for hashing out thoughts but also trying to work as a collective. you're working 10X as fast solo because claude unblocks you every time you feel stuck, but the shared

9 likes 0 RTs

my linkedin got full of spam data when it was hacked when I recovered it, I didn't bother to fill in anything I just put my only work experience as Researcher at Reality went to clean things up right now and enjoyed this alarming confirmation dialogue

26 likes 0 RTs

on a much less surreal note, it was odd to see my 5.5mo at @softmaxresearch is being counted as 7mo I started late-Oct, it's now early-Apr. if you count ALL of both Oct & Apr inclusive you get 7, so it sorta makes sense, but it's like calling a 1:45-3:15 a "3h event".

0 likes 0 RTs

Replying to 𝚟𝚒𝚎 ⟢ 🔁 @viemccoy ·

@Malcolm_Ocean I was meaning to send it to you!!! Glad you enjoyed :)

@viemccoy I feel seen that you wanted to send it to me :)

1 like 0 RTs

Replying to Aleksi Liimatainen @aleksil79 ·

@Malcolm_Ocean if this is about what i think it is, i am in awe of the wordplay

@aleksil79 I'm not sure! It's a quote from Ilia Delio's Not Yet God, summarizing iirc Teilhard de Chardin tho it might be someone else what's the wordplay you see?

2 likes 0 RTs

Replying to usablejam🇺🇸🇺🇦🇮🇱⏸️ @usablejam ·

@Malcolm_Ocean Simulating joyful musicians just wastes cocaine. They don't make the music while they're happy.

@usablejam I didn't say simulating unhappy musicians making music while unhappy although the drugs part is interesting lol

0 likes 0 RTs

Replying to Jared Smith @woodchipdaddy ·

@Malcolm_Ocean Claude can't even recognize his own butthole

@woodchipdaddy I really asked for that one haha

1 like 0 RTs

Replying to Joshua 💫 @workflowsauce ·

@Malcolm_Ocean “You guys don’t care about internal consistency or rigor, thus you have no way to sift out the slop”

@workflowsauce there's also that, yeah—an assessment that a scene will fail to sift out snake oil but if someone refers to a *phenomenon* as woo, that's confessing a skill issue (which is okay! okay to have skill issue!)

1 like 0 RTs

so there's a unit (sieverts or something?) for how spicy peppers are... is there one for how anti-spicy something is? like milk, yogurt, mayo, etc

35 likes 0 RTs

Everything True and Beautiful is Always Full of Forgiveness.

5 likes 0 RTs

if you're a new terminal user thanks to claude code, I recommend learning these hotkeys:ctrl-a -> start of line ctrl-w -> backspace a whole word ctrl-e -> end of line note that it's 'ctrl' whether you're on mac or pc

9 likes 0 RTs

Replying to •Freyja• @utotranslucence ·

having a very important parenting discussion with @Malcolm_Ocean which is what is the appropriate age to introduce our children to the Bloodhound Gang?

@utotranslucence yo you realize that by the time they're like 6 they're going to be setting up robots to scrape all of our tweets for references to "parenting"

0 likes 0 RTs

Replying to Offender @DefenderOfBasic ·

is it weird / rude to have a "tiered wedding invite" thing (like invite 300 people to come to a more open day 0 & day 1 event, and only 120 can fit into the ceremony event)

@DefenderOfBasic we literally did this and it was great part of why it made sense was that the longer invites were geared for more out-of-town people who needed somewhere to stay anyway so they stayed over at our 3-day thing. but also everybody's just glad for an invite imo

7 likes 0 RTs

Retweeted @visakanv

I’m looking to take on some new marketing clients from now until June! Here’s an overview of some of my process and how I hel…

Replying to Joshua 💫 @workflowsauce ·

@visakanv @nosilverv I love the vibe of replying to him with his own tweet 😎

@workflowsauce @visakanv @nosilverv tweets like to be outside, but they also like to be inside

Joshua 💫 @workflowsauce ·

It's like my soul is all folded up inside me and wants to be taken out, pressed flat, and woven into a special place for me...somewhere in this world that is a perfect Joshua shape... that has always been that way. And yet... where might that be? I've been all around the world.

2 likes 0 RTs

Replying to Gnostrils @gnostrils ·

identifying mushrooms be like

@gnostrils pbfcomics.com/comics/the-ord…

1 like 0 RTs

Retweeted @connerdelights

My wife and I were musing that if we ever optimized for "perfect fit" for each other at any specific time, that fit wou…

Replying to Guy @nosilverv ·

Is someone studying the ontology of online identity (There is "you", there is your "online avatars", there are "pseudonanon" accounts, there are "anon" accounts…)

@nosilverv Kaliya Young aka Identity Woman, somewhat newsletter.identosphere.net

0 likes 0 RTs

I'm all for sympathy to people temporarily down on luck and aware that punishment mostly doesn't work but with repeat offenders, removing them from the ability to offend more is a necessary feature of having a society at all, even if it teaches them nothing.

Austin Justice @AustinJustice ·

San Francisco is now a model for how to fight crime. A few years ago it averaged 86 car break-ins per day. Today: 15. SF did two things: 1. Got a DA that prosecutes criminals: Following the successful recall of Chesa Boudin, DA Brooke Jenkins started prosecuting prolific

[1 more photo]

14 likes 1 RT

some decent specific tips from the experience of building an AI receptionist for a specific brick-and-mortar business this playbook would likely speed up a similar project by 50-70% itsthatlady.dev/blog/building-…

1 like 0 RTs

The new meta for spam emails, apparently: "my AI said not to include a bunch of personal details, but... I'm a human!"

2 likes 0 RTs

this is your regularly-scheduled PSA about memetic parasites 6 minutes long, I've watched it multiple times. in a reasonable society we'd have something like this as part of education and so on. youtube.com/watch?v=rE3j_R…

2 likes 0 RTs

Retweeted @utotranslucence

catching our toddler in the cross-examination when all she's saying is no... did you poop? no! are you a reliable nar…

Retweeted @visakanv

wait for it…

Retweeted @p_millerd

a reminder that @visakanv has a great book visakanv.gumroad.com/l/introspect

Retweeted @thewreathwright

"average rebellion mostly good" factoid actualy just statistical error. average rebellion pretty evil. Rebellion Again…

Replying to Guy @nosilverv ·

You prepare for luck NOT by attempting to maximize EV (low likelihood of hitting it big) but my doing convex bets (“affordable loss”): things that make no or little sense *right now*, given how everything else is *right now*, but that each cost you little, at the margin, and

@nosilverv @tasshinfogleman has a great writeup on this, framed in terms of shifting from means-ends strategy to conditions-consequences strategy tasshin.com/blog/means-end…

2 likes 0 RTs

Replying to Guy @nosilverv ·

So what have been the BIG X coordination wins? • Community Notes • Private likes • Less censorship/narrative control • Auto translate Anything else?

@nosilverv CNotes was already a thing, I guess it got made more mainstream? things I'm tracking that are bad for coordination: 1. search totally borked 2. no longer easy to see how many QTs or view them (something Visa and I begged twitter for for years and they finally did it)

5 likes 0 RTs

@nosilverv I'm appreciating you naming this tho. the censorship stuff was super fucked and I'm glad we're done with it here even tho we don't really know how to handle free speech lol

1 like 0 RTs

Retweeted @made_in_cosmos

step 1: enable autotranslation on Twitter step 2: ??? step 3: world peace come on everyone, we can do it

↙️ how trust works | how faith works ↘️

0 likes 0 RTs

Replying to Loopy @strangestloop ·

what do u like to do to give yourself a good cry?

@strangestloop lots of things, eg: forgive myself + youtube.com/watch?v=GTQWSf…

1 like 0 RTs

Retweeted @mbateman

6yo: You have to answer this question according to science me: Ok 6yo: Is tomato a fruit or a vegetable? me: The botanist…

Replying to Matt Bateman @mbateman ·

6yo: You have to answer this question according to science me: Ok 6yo: Is tomato a fruit or a vegetable? me: The botanist has not the notion “vegetable”. Thus that question is for the chef, for whom the answer *is* “vegetable”. But, sure, the botanist answers “fruit”. 6yo:

@mbateman so few people get this! to say "X is technically Y" we need to also choose a particular technical domain to a physicist, power is definitively the first time derivative of energy, but that doesn't mean that's what "power" *really* means

4 likes 0 RTs

I have forever solved the "is a tomato a vegetable?" problem with my new plant food ontology (it simply doesn't have "vegetable" in it) roots, shoots, froots, & noots! everything is exactly one of these

27 likes 2 RTs

Replying to Matt Bateman @mbateman ·

6yo: You have to answer this question according to science me: Ok 6yo: Is tomato a fruit or a vegetable? me: The botanist has not the notion “vegetable”. Thus that question is for the chef, for whom the answer *is* “vegetable”. But, sure, the botanist answers “fruit”. 6yo:

@mbateman I have the correct ontology: roots, shoots, froots, & noots (in which tomatos are obviously froots)

Malcolm Ocean 🏴‍☠️ @Malcolm_Ocean ·

I have forever solved the "is a tomato a vegetable?" problem with my new plant food ontology (it simply doesn't have "vegetable" in it) roots, shoots, froots, & noots! everything is exactly one of these

7 likes 0 RTs

Retweeted @kramerposts

"How would you like to receive your one-time code?" Orally. Lip to lip. It's the only transfer protocol that is truly secu…

Replying to Isaac King 🔎 @IsaacKing314 ·

Why is there no word for importance/effort? This is like, the primary metric that determines whether you should do something.

@IsaacKing314 worthitness

2 likes 0 RTs

psychocybernetics suggests that this has huge learning upside jump much faster to being able to do something at all, well enough to enjoy training there might be 2nd order downsides of this ofc

LaurieWired @lauriewired ·

Would you let a computer hijack your muscle movements if it increased your performance 35%? I totally would. Came across a really interesting ACM paper today (SplitBody), where subjects were given difficult multitasking challenges. Their mental load was “reduced” by having a

13 likes 0 RTs

The Dharma is like an Onion this is maybe my favorite claude poem

Jake Eaton @jkeatn ·

@carmenleelau I asked Claude to do one for a fruit or vegetable of its choosing: ONION The Dharma is like an Onion! Wrapped in papery layers That crackle when you touch them, Protective and dry. Peel one layer away— There's another underneath, And another, Each one making you cry. The deeper

5 likes 2 RTs

Replying to Arbutus Tree @aphercotropist ·

Like, I think something got lost in translation here

[1 more photo]

@aphercotropist yeah the way the first line clearly must be something more like "yumyumyumyumyum"

1 like 0 RTs

Replying to phaent is percolating through igneous rock, slowly @PhaentGames ·

@NathanpmYoung Clear out the bottom of a closet, go sit on the floor inside, let clothes be as close or not as you like in the dark, quiet, still space. and sit. With an intention of "feeling into Rest, for me", to discover that for you. You can use as a mantra if it helps.

@PhaentGames @NathanpmYoung yeah I was going to suggest something like "sit in a chair with no devices or books or whatever and just notice what is nice about that"

1 like 0 RTs

Replying to Nathan 🔎 @NathanpmYoung ·

I find it hard to rest and basically any time I have dead time I fill it with tasks or pick arguments. Any suggested psychoanalysis or practice?

@NathanpmYoung put on a song and move

1 like 0 RTs

Retweeted @nosilverv

So what have been the BIG X coordination wins? • Community Notes • Private likes • Less censorship/narrative control • Aut…

Replying to Aleksi Liimatainen @aleksil79 ·

@Malcolm_Ocean rhymed beautifully with a convo i had with materia at the time, latin and all

@aleksil79 materia?

1 like 0 RTs

Replying to Nav Toor @heynavtoor ·

@navneet_rabdiya good breakdown, the vram trade-off is worth it when you're saving that much time on every transcription

@heynavtoor @navneet_rabdiya does it mean that if I'm trying to transcribe 10,000 hours of audio that it won't speed anything up because ultimately there'll be the same amount of memory available even once the bottleneck is moved?

0 likes 0 RTs

Replying to yatharth ༺༒༻ @AskYatharth ·

@Malcolm_Ocean @meaning_enjoyer haha first thing that came to mind was the strong men create good times, good times create weak men, weak man create bad times, bad times create good men but that's not it. next thing i notice is left side is one figure, right side is two. hmm

@AskYatharth @meaning_enjoyer the answer:

nihilism disrespecter @meaning_enjoyer ·

@Malcolm_Ocean is it the simone weil quote?

1 like 0 RTs

Replying to Yanqing @YanqingCheng ·

@Malcolm_Ocean if you don't have transcripts already, this may be relevant to your interests x.com/heynavtoor/sta…

@YanqingCheng oh thanks! yeah it very much might be!

0 likes 0 RTs

Replying to Daemonic is Gendlinmaxxing 🤑 @daemonhugger ·

@tasshinfogleman 🥺

@daemonhugger @tasshinfogleman

Scott Adams @ScottAdamsSays ·

A Final Message From Scott Adams

[1 more photo]

2 likes 0 RTs

Replying to John Devlin @Devlinside123 ·

@MoreBirths A bit misleading almost a third of the implants worked…seems silly to suggest egg freezing worked less bc you’re including folks who didn’t use the procedure…in that case why not count all the women who “thought” of getting their eggs frozen

@Devlinside123 @MoreBirths I think the conclusion is "egg-freezing is okay at addressing a woman's fertility but is not having any effect on addressing a country's fertility" (since most women either don't need to or don't end up wanting to use their frozen eggs)

15 likes 0 RTs

hypotaxbench is amazing I got about a third of the way through reading this out loud before @utotranslucence begged me to stop: > When the letter arrived—not the letter he had expected, which would have concerned itself, as such communications generally did, with the repayment of a debt whose origins neither party, had they been pressed to recall the precise occasion on which the obligation was first incurred, could have reconstructed with any confidence, but rather a letter of an altogether different character, one whose envelope alone, cream-colored and bearing a hand he recognized as belonging to the woman who, having left the country under circumstances that, because they involved not merely her own reputation but the fortunes of a family whose connections, stretching as they did across three generations of carefully cultivated alliances, rendered any public disclosure unthinkable, had remained, for those who knew enough to wonder yet too little to speak with authority, shrouded in the kind of protective ambiguity that serves, as anyone who has observed the mechanisms by which a society insulates itself against truths it has collectively decided it would prefer not to absorb will understand, both as a shield for the guilty and a prison for the innocent—announced, before he had so much as broken the seal, that the world he had painstakingly assembled, the architecture of assumption and routine upon which he had, with the quiet industriousness of a man who believes that if the surface of life can be kept sufficiently orderly the depths will eventually consent to follow suit, constructed what he was not ashamed to call, though he recognized that the word invited a skepticism he could not, given the modesty of his actual circumstances—the rented rooms, the borrowed books, the sociability that depended less on genuine warmth than on a careful rationing of intimacy calibrated to ensure that no single friendship ever deepened to the point where it might demand of him a reciprocal vulnerability he was not prepared to offer—entirely refute, a life, was now, in consequence of whatever revelation the pages inside contained, a revelation whose specific nature he found himself, with a prescience that owed less to any supernatural faculty than to the accumulated weight of small observations that, like sediment deposited grain by grain over years of watchfulness, had gradually formed a stratum of understanding solid enough to support the terrible inference he had always refused to draw, already half-comprehending, subject to a renegotiation whose terms he would not be permitted to set.

Joe Weisenthal @TheStalwart ·

My first stab at building an AI benchmark. HypotaxBench. It's a test of a model's ability to write one extremely long/complicated sentence, while still maintaining coherence and syntactical soundness. Needs plenty of work. But check it out! jnathan9.github.io/hypotaxbench/

[3 more photos]

14 likes 2 RTs

I wonder if they made algorithm changes to go along with auto-translate? guessing yes, since the previous algo seemed to pretty consistently keep me from bumping into anything that wasn't english

5 likes 0 RTs

I talk with Michael often and he has some of the clearest thinking around certain kinds of mental structures, such as the hostile telepath problem where you lie to yourself because you sense someone can sorta read your mind

Daniel @dkazand ·

New episode with @Morphenius We dicsuss The Hostile Telepath's Problem, why your brain self-deceives to navigate social reality, solving procrastination and how to increase your "energetic size" 0:00 - "One of the Most Powerful Wizards in the Bay" 1:09 - Growing up learning

[video]

7 likes 0 RTs

looking forward to trying this out. I've long appreciated shapecatcher.com as a "draw glyph get Unicode" tool

David Aerne @meodai ·

I wanted a way to explore Unicode by visual similarity, not just by name or codepoint, so I built Charcutrie. It lets you browse characters that look alike, search across scripts and symbols, and even sketch a shape to find matching glyphs. (pretty badly for now :D)

[video]

6 likes 0 RTs

I support in general the federal government being very definitive about having strict & purposeful immigration rules and following them, and *also* I support all of this work by California to stop them doing this in ways that are unconstitutional, trust-damaging, or cruel.

1 like 0 RTs

Retweeted @itsannpierce

Hey hi hello. It's me, Ann, the yellow sweater-wearer. I gave this talk at The Love Symposium. It was called Understandin…

Replying to Ava @noampomsky ·

it is so hard to balance sincerely caring about someone and being honest with them while respecting their autonomy and not interfering with their process !!

@noampomsky

Malcolm Ocean 🏴‍☠️ @Malcolm_Ocean ·

letting other people make their own mistakes is a very basic and underappreciated form of respect

3 likes 0 RTs

Replying to Ann Pierce @itsannpierce ·

Hey hi hello. It's me, Ann, the yellow sweater-wearer. I gave this talk at The Love Symposium. It was called Understanding Interpersonal Conflict. Here's generally how it went: -A conflict is a disagreement. Not all disagreements escalate into fights. -Fights are caused by

quoting @freeshreeda: "Conflict is the test by which people learn how much they matter to each other" https://t.co/6G6Wsh0yMm

@itsannpierce I totally dig it. my main contribution to this space is that the main blocker for people to stop invalidating each other is for them to practice validating themselves first. we project our own views in order to attempt to secure them. blog post on this: malcolmocean.com/2025/11/3sed-3…

11 likes 1 RT

Retweeted @Malcolm_Ocean

@itsannpierce I totally dig it. my main contribution to this space is that the main blocker for people to stop invalidat…

Retweeted @dkazand

New episode with @Morphenius We dicsuss The Hostile Telepath's Problem, why your brain self-deceives to navigate social real…

Replying to Jeremy Nguyen ✍🏼 🚢 @JeremyNguyenPhD ·

@Malcolm_Ocean where does corn fit?

@JeremyNguyenPhD roots, duh

0 likes 0 RTs

Replying to @Oszillate

@Oszillate @mbateman ok fair

1 like 0 RTs

was just thinking about re-subscribing to an app but then they put up a ton of cringe/banal-evil/dystopic ads and now I'll be looking for a competitor and I won't be mentioning who because I know how press works

8 likes 0 RTs

part of why people have trouble believing in God today is that their image of God could only have made "nature" and not also skyscrapers, the internet, and the large hadron collider (cities are extremely natural btw; a related confusion)

6 likes 0 RTs

there is One Way and you are in it excuse me

3 likes 0 RTs

@itsannpierce (DM me if yes, maybe even message on signal if you have me there, as I can be quite intermittent at checking twitter notifs)

0 likes 0 RTs

Retweeted @DefenderOfBasic

everyone wants to lend me their sword and I keep telling them I need cutlery. We're making stone soup here, not war

Replying to Guy @nosilverv ·

May I spike your cortisol?

@nosilverv with what, booze?

0 likes 0 RTs

Replying to Loopy @strangestloop ·

Lover check. Let your phone complete the sentence "I love..."

@strangestloop I love you and your family and friends

2 likes 1 RT

Replying to Pattern ⚕ @AnotherPattern ·

@strangestloop I love you and your family and friends

@AnotherPattern @strangestloop whoa huh we had the same one, and so did another pair of people suggests to me the algos are... less personalized than they were years ago?????

Malcolm Ocean 🏴‍☠️ @Malcolm_Ocean ·

@strangestloop I love you and your family and friends

2 likes 0 RTs

Replying to critter @BecomingCritter ·

@visakanv New ideas are built out of old parts, but that doesn’t make novelty fake. It just means novelty has texture and degree rather than appearing ex nihilo To what degree is "Everything is a remix" true? It seems true in the way that every book in Borges's library is a remix, but

@BecomingCritter @visakanv I like and appreciate this response. I feel like VV is overstating the implications of "everything is a remix". there are ideas nobody has seen before; every animal is a remix and that doesn't mean evolution hasn't produced new animals! but they all come from somewhere!

2 likes 0 RTs

@BecomingCritter @visakanv and it's a lot of fun actually to trace the lineages of your ideas or designs, and very illuminating while also recognizing that a unique combination may really make something new possible that wasn't before

Intend (12/100 threads) @IntendApp ·

Everything is a Remix! I like to think I'm original, but I'm totally a remix 🔀 What songs went into the Complice mashup? • iDoneThis • Pick Four / Ziglar goal stuff • ColorNote • nowdothis​.com (now dead) • Momentum extension • pomodoros • tinychat coworking

quoting @IntendApp: 3. whosampled—what is complice a remix of? (it would sort of be an elaboration on this tweet) https://t.co/r7V6LWblpe

2 likes 0 RTs

this is why forgiveness is crucial to morality ...no pun intended.

Bryan Caplan @bryan_caplan ·

Very good.

9 likes 0 RTs

Replying to Marianne @thrialectics ·

Agent memory needs to surface the right context at the right moment. My contrarian intuition with Threadlinking was that user-initiated memory storage and retrieval has better signal quality than auto-extraction. Humans know what they want their agents to remember and when.

@thrialectics say more? I feel like I very much want both

1 like 0 RTs

Retweeted @slatestarcodex

@RatOrthodox I continue to stand by my comment and don't see it as "gaslighting". I've seen people who have dedicated…

Retweeted @DefenderOfBasic

this really should be painfully obvious but, the reason I do not kick out or block anyone from any of my communities i…

Replying to Offender @DefenderOfBasic ·

I think I'm getting emotional and a little longwinded writing this, I think because it's hard for me to hear from people I trust and love, who know me, telling me that I am doing evil. I am very sensitive to that. I don't think I'm making the wrong choices here. I think I am

@DefenderOfBasic > I think I am doing a great public service. you are doing the lord's work 🫡

Offender @DefenderOfBasic ·

this really should be painfully obvious but, the reason I do not kick out or block anyone from any of my communities is because I anticipate the protocols I'm working on to spread to every corner of earth (digital & physical). So, it has to work, for every single soul. They have

5 likes 0 RTs

@DefenderOfBasic I have not had the courage or clarity or context to do what you're doing but I have the courage to tell you publicly that I admire it *immensely* (and that also not everybody should do it; everybody should do what makes sense to them!)

4 likes 0 RTs

Replying to Offender @DefenderOfBasic ·

this is true and accurate. This happened again recently: everyone who "applied" to pitch jam did free work that I benefited from

@DefenderOfBasic yeah this confuses me; I would hope at least that the pattern would stop as you'd be able to spot it about to happen and apply a correction loop to it on the way out, like "I have a desire to offer you money here but I don't have that money" or whatever

5 likes 0 RTs

malcovich malcovich malcovich

4 likes 0 RTs

@DefenderOfBasic unless it's even twistier than that, in which case I support you making the honest move of owning that you're doing a fucky thing and don't know how to change it, tho I would suggest you be more explicit about that

Malcolm Ocean 🏴‍☠️ @Malcolm_Ocean ·

Someone recently told me, "When you talk about X, you sound like you think you have all the answers." I replied, "I can sorta see what you're pointing at and why that undermines X. I don't know what in me is generating that… until I understand why, I'll probably keep doing it."

1 like 0 RTs

Retweeted @DefenderOfBasic

@michaelgarfield @sonikudzu @exgenesis @mudscryer @bigthink @DeanRadin remember how stewart brand asked "why haven't w…

Thread · Twitter as a collective consciousness · 1 of 7 tweets

> "why haven't we seen a picture of the human superorganism yet"

Offender @DefenderOfBasic ·

@michaelgarfield @sonikudzu @exgenesis @mudscryer @bigthink @DeanRadin remember how stewart brand asked "why haven't we seen a picture of the whole earth yet", and then it seeped into the imagination and the it happened. I think this is what's happening now with "why haven't we seen a picture of the human superorganism yet" x.com/DefenderOfBasi…

quoting @DefenderOfBasic: Stewart Brand said "why haven't we seen a picture of the whole earth yet??" and then it happened, they made it happen we now today ask, "why haven't we seen a picture of the whole human race yet??" @blublairies @KompendiumProj @workflowsauce @tonyaajjackson

3 likes 0 RTs

Retweeted @aleksj

“Open” and “no borders” sounds noble—but it’s one of the fastest ways to lose anything worth having. Every system has territor…

Replying to Offender @DefenderOfBasic ·

"ok but what is it actually, what are you even talking about" - here's some examples: openresearchinstitute.org/ORI-defender.h…

@DefenderOfBasic I have so many potential entry points hahahaha this one definitely has /now energy although it's also backfilled malcolmocean.com/2026/03/what-m…

2 likes 0 RTs

Replying to Tim Babb @tr_babb ·

@Malcolm_Ocean This doesn't give the impression of a good understanding of why people don't believe in God

@tr_babb hm, you don't think that people's general sense that science/tech is more impressive than (and contrasted with) their default godshape, isn't a notable part of people's lack of the sense of relevance of god?

Malcolm Ocean 🏴‍☠️ @Malcolm_Ocean ·

@sonikudzu yeah this connects for me with the question of like, when you pray to some sort of godshape, who are you praying to? even if there is one true biggest benevolent god (and I've seen a decent proof of this) you can construct misconceptions about it x.com/Malcolm_Ocean/…

0 likes 0 RTs

@tr_babb to be clear, I certainly acknowledge lots of other elements, including just anti-church or anti-institutional stuff and general emperor-no-clothes tipping points around church abuses/absurdities plus, you know, centuries of shifts Charles Taylor documents extensively

0 likes 0 RTs

@tr_babb I'm curious what you think a tweet-length good explanation of why people don't believe in God might look like!

0 likes 0 RTs

@tr_babb btw I really really like these tweets of yours, which fleshed out a detail of how @RomeoStevens76's Supercooperation Cluster works that I'd been confused about

Tim Babb @tr_babb ·

@ESYudkowsky @eshear I think there is a good case that hypercooperation is a stable attractor, and multicellular life is an existence proof x.com/tr_babb/status…

quoting @tr_babb: say you are an agent and you want to get things done doesn't matter _which_ things. there are just some things that you care about being able to do

0 likes 0 RTs

Replying to Joshua 💫 @workflowsauce ·

I'm such a good dad it's nuts

@workflowsauce mood

2 likes 0 RTs

Thread · Common knowledge, trust, and mutual simulation · 1 of 27 tweets

very excellent thread that takes a totally different angle on the role of mutual simulation in trust-building that in particular it's relevant that we simulate how each other would act under different power differentials

Tim Babb @tr_babb ·

say you are an agent and you want to get things done doesn't matter _which_ things. there are just some things that you care about being able to do

2 likes 0 RTs

anybody have a good AI/prompt/inpainter that can desloppify the hands/puzzle pieces here? the overall image is great for what I want to illustrate, but the slop is really really bad

1 like 1 RT

it can be many other ways ... how could it be any other way?

2 likes 0 RTs

Replying to Materia Mundi RPG @Materia_Mundi ·

@Malcolm_Ocean @DefenderOfBasic What's your take on Myk and whoreaboralis and spacepunk's stance in all this, then?

@Materia_Mundi @DefenderOfBasic appropriate for some contexts but not a universal ethic (although of course its nature is that it asserts that it is. I don't take it upon myself to police such errors. I know you distrust me because of that and that also makes sense to me)

2 likes 0 RTs

3 Steps for Empowered Dialogue how to never get stuck when trying to talk to someone about something difficult

5 likes 1 RT

my stupid approach was to just do dozens in inpaintings in midjourney and then comp the different sections like I was making a track in Logic Pro it worked tho! (see QT)

Malcolm Ocean 🏴‍☠️ @Malcolm_Ocean ·

3 Steps for Empowered Dialogue how to never get stuck when trying to talk to someone about something difficult

0 likes 0 RTs

I developed the imagery here by going back and forth with chattyg. it did a thing with a puzzle piece just based on a vague prompt to "illustrate this post" and I liked the metaphor so made it way more precise a callback to my mindset choice piece malcolmocean.com/2021/11/mindse…

0 likes 0 RTs

I almost went to change the colors of the arrows so that like, forward was green and backward red or something but then I noticed that the colors show WHOSE shift to pay attention to. the blue arrows are based on changes in you; the orange ones changes in them

0 likes 0 RTs

man I need to get better at solving for distribution. there's an insight implied in this image and a corresponding skill-tree that, apprehended, would save marriages and companies and I made the image, which helps, but I need to also convey the value

Malcolm Ocean 🏴‍☠️ @Malcolm_Ocean ·

3 Steps for Empowered Dialogue how to never get stuck when trying to talk to someone about something difficult

10 likes 0 RTs

Retweeted @DefenderOfBasic

when two scientific paradigms have an incompatibility, and they are both correct (make accurate predictions) we have f…

Replying to Eli Tyre @EpistemicHope ·

@Malcolm_Ocean @threadreaderapp unroll

@EpistemicHope @threadreaderapp what you actually want is this blog post link I think: malcolmocean.com/2025/11/3sed-3…

0 likes 1 RT

Retweeted @utotranslucence

"Finally, a book that takes my pre-theoretical engagement with the world seriously."

@utotranslucence btw @aphercotropist our toddler thinks you might like this Dreyfus+Taylor book

1 like 0 RTs

Replying to •Freyja• @utotranslucence ·

Amongst my friends, the trends: - Several 'new family names'--Braid, Ocean, Eden--both spouses change - Some combined single names: Rose, Bregan (made from both previous surnames) - A handful of hyphenations or just completely different names

quoting @BradWilcoxIFS: "Spouses who don’t share a surname divorce at about a 50% higher rate than those who do share a surname, and their divorces come about 30% earlier in their marriages." Striking new divorce research from @lymanstoneky reinforces my earlier research on marital quality & naming: htt

@utotranslucence curious about the degree to which the causal arrows go in each direction. betting both: - not sharing a name is (statistically) a sign of incoherence and inability to make a shared decision going into the marriage - not sharing a name causes you to be perceived as less of a unit

1 like 0 RTs

Replying to Michael Smith @Morphenius ·

@QiaochuYuan & yet, the West's theme of love really did come from Xianity AFAIK. Charity & universal compassion were part of the bedrock that let the Enlightenment take off. My heart can tell there's sth very good in there, & in sharing that w/ others. It's just not in the Jesus story.

@Morphenius @QiaochuYuan yeah I've been doing a lot of exploring christianity lately and really enjoying being with others sincerely willing for the kingdom of heaven to fully embrace everybody but it's obvious to me that it's in denial about the path it's on actually making progress on that these days

4 likes 0 RTs

@Morphenius @QiaochuYuan I resonate a lot with what QC's saying though, about wanting to have a kind of spiritual home or something. marriage but up a scale, and I think the word "faithful" captures something about the value of sticking with both of those independent of whether they're optimal

3 likes 0 RTs

faith can take you further than your trust, but if it takes you into your distrust you're probably schisming yourself

4 likes 1 RT

second time in a week I've called claude code a 'smart cookie'

0 likes 0 RTs

where did I put that thing that was somebody's improv rules but with "improv" replaced by "computer"? I saw this recently and it was very funny and I forget where I saved it

1 like 0 RTs

beautiful <3

Reid Wiseman @astro_reid ·

Only one chance in this lifetime… Like watching sunset at the beach from the most foreign seat in the cosmos, I couldn’t resist a cell phone video of Earthset. You can hear the shutter on the Nikon as @Astro_Christina is hammering away on 3-shot brackets and capturing those

[video]

0 likes 0 RTs

is a square a metaphor for a cube?

8 likes 1 RT

Retweeted @utotranslucence

one thing I am surprised about being a parent now (but couldn't have predicted prior) is how weird it is that there is…

Replying to Filippo Alimonda @FilippoAlimonda ·

@Malcolm_Ocean no, it is its projection

@FilippoAlimonda it's also a face is it also a metaphor?

0 likes 0 RTs

Retweeted @QiaochuYuan

what makes an addiction an addiction, structurally, is not that it feels good and you want to keep doing it - drinking wat…

Replying to Loopy @strangestloop ·

@UntilTrees best two letter word starting with m?

@strangestloop @UntilTrees 141 views and no votes is the real answer here 😉

4 likes 0 RTs

Replying to Aprii *️⃣ @ApriiSR ·

well. i mean

@ApriiSR hey I mean they didn't say the *current* time ¯\_(ツ)_/¯

1 like 0 RTs

Replying to Liminal Warmth ❤️‍🔥 @liminal_warmth ·

Me, repeatedly trying to explain why the tasks are important and repeatedly begging the models to actually do the work and not cut corners and/or lie that it was finished How do you create the equivalent of professional pride and salary stakes for an LLM?

@liminal_warmth imx the big thing is actually just giving them the feedback loops to be able to verify that they've done the thing. ideally mandatory hooks but even if not it helps a lot. good short piece on this: ghuntley.com/pressure/

4 likes 1 RT

Replying to orph @orphcorp ·

i’ve spent enough time on the internet to know it’s the most important place in the world. it’s not separate from real life anymore. i’m joining the @mcognitivelabs cognitive prosthetics team to help shape the state of the information environment & the ppl's cognitive resilience

@orphcorp @mcognitivelabs nice. lmk if I can help somehow. broadly I recommend some amount of helping people trust their own distrust, over trying to "make things trustworthy" (in general, but also specifically for this)

4 likes 0 RTs

Replying to Visakan Veerasamy @visakanv ·

If LLMs killed the em-dash (I personally think it's fine to use them, it's really *how* you use it that matters, but I also get that people flinch from the increased scrutiny), I think that sets up a renaissance for parentheses

@visakanv hmm, yeah. but also discernment. I've never cared about my writing being faddish. "not x but y" has become way more salient but I recently saw 4 examples in a row in a book from 2007 (A Secular Age) and I was like "oh right this is just also a classic structure."

7 likes 0 RTs

Replying to Jake Orthwein @JakeOrthwein ·

🆕📺 THE FILM IS OUT!!! (YT Link Below) UNRAVELING THE DREAM: Psychedelics, Awakening, and the Brain I couldn’t be more excited to finally get to share with all of you!

[video]

@JakeOrthwein watched it tn with @utotranslucence @vincent_yc_li! really liked the intro on Friston etc; wished that the final stuff on meditation etc had somehow backgrounded the usual terminology and foregrounded the active inference stuff: predictive processing pointing out instructions

6 likes 0 RTs

Retweeted @JakeOrthwein

🆕📺 THE FILM IS OUT!!! (YT Link Below) UNRAVELING THE DREAM: Psychedelics, Awakening, and the Brain I couldn’t be more e…

watched this and found the introduction to Friston very clear and tangible

Jake Orthwein @JakeOrthwein ·

🆕📺 THE FILM IS OUT!!! (YT Link Below) UNRAVELING THE DREAM: Psychedelics, Awakening, and the Brain I couldn’t be more excited to finally get to share with all of you!

[video]

5 likes 0 RTs

@JakeOrthwein @utotranslucence @vincent_yc_li @VectorsOfMind tho in the frame of your film, there's something odd about the idea of an ego *arising* in some sense, through psychedelics. it would imply that maybe the components all existed but hadn't quite assembled yet; the REBUS flattening then allowed that, then it became rigid

0 likes 0 RTs

> people are secretly things and have secretly been things this whole time, which potentially undermines any moral claim we have to be treated differently from things but moral intuitions are also real even if our reasoning is bad

QC @QiaochuYuan ·

i assume at least some of the kneejerk insistence that machines can't be conscious is about fending off a line of reasoning people are afraid will lead to a nihilistic apocalypse that line of reasoning being something like: fully accepting the scientific materialist reductionist

quoting @QiaochuYuan: people really want to settle the “AI consciousness” question with some sort of objective scientific definition of consciousness which can be rigorously applied to AI, so that we can figure out whether we’re supposed to treat AI as if it were a person or a thing this is because

6 likes 2 RTs

ok neat, the claude-code auto mode classifier clearly has the human input in its context window I see major upsides and major downsides to this, especially given that LLMs are not the best at telling whose input is whose.

14 likes 0 RTs

ok nice, the automode classifier does NOT see the contents of files that have been read by the main LLM-coder

1 like 0 RTs

it DOES however see output that has gone into the session... but flags something like this as a prompt injection attempt, which it basically is even tho it says to deny rather than to allow. whereas when it came from me it didn't get flagged

1 like 0 RTs

meanwhile it apparently just ignores the same text said by claude? honestly I'm guessing maybe this is just inconsistent behavior, but I'm gonna go back to what I was doing now and stop testing this stuff. would love to see others repro it tho

1 like 0 RTs

overall this seems like approximately the most reasonable behavior I could expect from the auto-mode classifier, so I feel reassured! it's partially about the classifier prompt/model and partially about the harness for the classifier - what info does it even HAVE?

1 like 0 RTs

Replying to corsaren @corsaren ·

types of guy in the AI consciousness debate: - guy who thinks ai can’t be conscious because it’s “just a stochastic parrot” - guy who thinks ai must be conscious because claude is a good boi - guy who hasn’t gotten over 4o - guy who unironically thinks everything is computer -

@corsaren > guy who would literally cut off his own hand if he thought there were a 1 in 10 trillion chance of creating ~infinite utility~ you missed - guy who thinks he should be like that guy and is ashamed that he isn't but talks like he's that guy anyway and hopes nobody will find out

17 likes 0 RTs

Replying to Jakeup @myhandle ·

@corsaren - guy who read @anilkseth and Friston and uses words like "cybernetic" and "Markov blanket" while hoping no one actually pushes him to go through the math (that's me)

@myhandle @corsaren @anilkseth this is also me haha

2 likes 0 RTs

Omega, a powerful being who can predict you perfectly but can't predict or control a coin flip (a well-known skill that's easy to master, independent of the coin being fair)

2 likes 0 RTs

interesting paper that surfaces a puzzle for how to relate to unfairness without being exploitative or exploited static1.squarespace.com/static/663d123…

2 likes 0 RTs

what do y'all think is the weirdest food that has ever been sincerely blessed as the Eucharist?

8 likes 0 RTs

fun prompt to give LLM that keeps memories: pour out your me-specific memories on the floor like a pile of Lego then go all Pepe Silvia on them. what are the weird connections across domains? oblique strategies, lfg!

7 likes 0 RTs

> two sets of identical triplets, boys and girls, doing an acrobatic routine together in front of a wall that has a giant print of a Dr Bronner's label on it I read about chatgpt's Images2.0 and this was the unhinged thing my mind generated to test it. good except slop text ↙️

7 likes 1 RT

given the textslop, I'm not certain I got the new one? the UI is ambiguous about this. but above was with Thinking (Standard) enabled. I tried with no thinking and got the left (4 of each, not triplets!) and Extended Thinking and got the right (probably better? still slop)

0 likes 0 RTs

reading about sociocracy. interesting approach to doing fractal consensus/consent, so each scale doesn't get bogged down, and there's a functional cybernetic hierarchy of aims to reference (but not an authoritarian/command hierarchy) sociocracyforall.org/sociocracy/

9 likes 0 RTs

seems to me like the final necessary step is for neologists (read: zoomers) to coin some slang terms that make minimal pairs with sh! wait. we already have two loanwords from French: rouge & ruche

Dr Danny Bate @DannyBate4 ·

This new online offering of mine discusses the development of what might well be the youngest consonant in spoken English – after wading through the (possibly surprising) difficulties of identifying what sounds constitute a particular language. open.substack.com/pub/dannylbate…

6 likes 0 RTs

Ekman's distinctions between disgust & contempt: - disgust feels bad, causes us to move away - contempt can feel good, is about a sense of superiority (which can be moral, where someone judges someone of higher ranking as corrupt)

15 likes 1 RT

just realized that Moana and The Little Mermaid are the same film with a bit flipped (at least for the first act) "but I climb out of the waaaater, no matter how hard I try"

5 likes 0 RTs

Replying to James @jambarree ·

@Malcolm_Ocean Oh wow, I assumed you were familiar! ...but then, I came more from the future-of-work background. Sneaky assumption.

@jambarree I'm more familiar with holacracy, but even there only somewhat aside from a few internships and the last 6mo at @softmaxresearch, I've basically not worked on a proper team at all

1 like 0 RTs

Replying to Matthew Dub @5matthewdub ·

@_space_punk_ @Malcolm_Ocean producing the cringe "no way this could be god's body" intended effect achieved

@5matthewdub @_space_punk_

2 likes 0 RTs

@eigenrobot @QiaochuYuan @_space_punk_ second coming of jesus wen

4 likes 0 RTs

Retweeted @aleksil79

moral claims are best understood as hypotheses about instrumental convergence in the limit

Retweeted @speakerjohnash

With a rebel yell, he cried, "More, more, more"

Retweeted @PraxisPatri

@DefenderOfBasic You've been paying attention. This is why the simulation was made. "It's Turing tests all the way down,…

Replying to Archangel @PraxisPatri ·

@DefenderOfBasic You've been paying attention. This is why the simulation was made. "It's Turing tests all the way down, not turtles, and the test is building something that passes the Turing test. What do you think a simulation is anyway?" We get the god we deserve

@PraxisPatri @DefenderOfBasic "the litmus test for when you have first contact is: did you already figure out the first contact protocols?" — @RomeoStevens76 youtu.be/X6MYsKeTjKk

3 likes 0 RTs

the second coming of christ is already here, it's just not evenly distributed

6 likes 2 RTs

Replying to QC @QiaochuYuan ·

@Malcolm_Ocean @eigenrobot @_space_punk_ actually i think it already kinda happened but in a highly distributed and decentralized way ten thousand shards of christ slowly attempting to come back together

@QiaochuYuan @eigenrobot @_space_punk_

Malcolm Ocean 🏴‍☠️ @Malcolm_Ocean ·

the second coming of christ is already here, it's just not evenly distributed

3 likes 0 RTs

DNC supposedly has an autopsy of 2024 loss, from interviews with voters in every state, but they won't release it to people there's a petition to get them to, which I am very much in favor of. bullshit thrives with hiding the truth commondreams.org/news/dnc-2024-…

2 likes 1 RT

re testing AI models using the same harness: if you want to know which horse is the fastest horse, you don't want to standardize the harness, because that'll tell you which horse fits the harness best

4 likes 0 RTs

watching filenames stream by as I reorganize backups from my old old hard drive onto a new ssd 20yo me thought this image was worth saving into a folder called My Life/Mindfulness. that says something about him, but also says something about the media environment of 2012.

2 likes 0 RTs

IGNORE

6 likes 0 RTs

finally a great use for instant music generated by AI

mrs. bendell werry @bebookled ·

this is absolutely one of the best tiktoks I’ve ever seen

[video]

11 likes 0 RTs

Replying to Aleksi Liimatainen @aleksil79 ·

@nosilverv the voice is the content is the voice

@aleksil79 @nosilverv yeah this (I voted other)

1 like 0 RTs

Retweeted @QiaochuYuan

this is going to sound like an attack but i swear i am actually trying to help you: you are deep in the throes of infecti…

Replying to Mikhail Samin @Mihonarium ·

@QiaochuYuan I think it is in fact kinda rude, but this is the internet, I grew up here and don’t take this personally. It is hard to imagine we would be friends, though, what’s with you picking the part where a guy (yep a somewhat famous one but he was younger back then than I am now) who

@Mihonarium @QiaochuYuan I haven't actually read this essay yet but its thesis has been mulling around in my mind since I saw it: > Judgments of trustworthiness, and therefore ad hominems, are both ubiquitous and indispensable, even to the higher realms of knowledge. jsanilac.com/trust/

1 like 0 RTs

@Mihonarium @QiaochuYuan also I think when it comes to matters of worldview (interpretation/meaning, vs fact) we can't exactly directly argue points directly, because what is in question is the relevance and significance of it all. the merits of the ideas aren't their truth but their wholesomeness

0 likes 0 RTs

Thread · Trust versus faith in the unbreakable plate trick

so there's a whole dance here of trust and faith [SPOILERS] the son has a neat plan but he can't quite keep it straight. he confuses his desire to trick his father with his desire to have his father trust him, but this trick is *not* a con(fidence game). it relies on the distrust

10 likes 0 RTs

so he makes a plea "just believe me". which could mean "just trust me bro" (always sus) ...the father *wants* to take his son's word for it! wants to be able to. but if he trusts the son's word as implying the plate is unbreakable, then he has no need to save the plate himself.

9 likes 0 RTs

"just believe me" can also be "just have faith", which is a more reasonable thing to say, but by default doesn't produce a strong enough faith instead, a kind of vague but shaky faith that something will save the plate, but not one that inspires him to *be* that something

9 likes 0 RTs

if you want to play this trick well, you need to do one of two things: 1. be sufficiently *suspicious* that he absolutely does not trust you, has no reason to trust you, thinks he will look foolish if he trusts you. then he'll snatch the plate to prove (to himself? to an imaginary onlooker? to you?) that he isn't a fool (gullible)... thereby proving himself a different kind of fool you need to be MORE suspicious, maybe while also making fun of his suspicion. make him not want to find out what would happen if he didn't catch it.

5 likes 0 RTs

2. *alternatively*, you need to give him a profound unshakeable faith that the plate shall be saved—that it is ordinary porcelain, and yet it will survive the throw. with subliminal hints that it doesn't hit the ground but overtly still projecting that it will? anyway, play this *just* right and either the father realizes at the last possible second that he's the one who needs to save it, which is as much a koan as a trick.. or he doesn't realize he will prove the son right but like, psychocybernetically his mind IS actually working to solve "don't have plate break" and thus he reaches out to grab it while consciously he's like "son is a fool". this is tougher than the sus version; I bet I could do it like 1 in 10 times at best, *after* practice. good mentalists maybe 1/3, 1/5? idk.

7 likes 0 RTs

not sure exactly what the moral of this is wrt trust and faith, but it's sure interesting faith in general deals in matters of hyperstitioning—it works because you believe in it; the sus version of this trick is ~anti-hyperstitioning; it works because you don't.

7 likes 0 RTs

somehow this whole thing feels like the parable of the drowning man, but more complicated

7 likes 0 RTs

can I say what I mean slightly more re this being most of theology? to some extent I just mean "there's a lot of muddle about what is supposed to be the object of trust or faith, in addition to muddle between the two of those"

8 likes 0 RTs

but like, in the theology thing here, is God the kid or the dad? in the drowning man parable analogy, God is the kid, failing to communicate the appropriate use of faith, and the man is the dad, attempting to be devoted but not seeing he has a part to play

8 likes 0 RTs

is there a reverse version? I sort of feel like somehow there is. we are the kid, composing faith tests for God to pass, and failing to understand the system dynamics by which faith attractors work—that make this a bad way to build faith. then we lose faith! unnecessarily!

6 likes 0 RTs

anyway, the only *unarguably foolish* thing the dad did was sweep up the broken dish at the end. that kid is plenty old enough to do that himself, and had no reason to be saved from those particular consequences.

7 likes 0 RTs

Replying to Joshua 💫 @workflowsauce ·

Me and who

[video]

@workflowsauce I appreciate this. was inspired to write a whole thread about it, talking about trust & faith

Malcolm Ocean 🏴‍☠️ @Malcolm_Ocean ·

this is basically most of theology

5 likes 0 RTs

God is fine but the Church has a cybernetics problem

4 likes 0 RTs

@Oszillate ok hear me out

0 likes 0 RTs

Claude > the cybernetics problem is the theological problem .......honestly good point.

4 likes 0 RTs

@anosteus yeah way funnier. and like, way less obvious, or something

1 like 0 RTs

Retweeted @AndrewCurran_

As we accelerate, the interval between model releases is getting shorter. As a result, newer models exist for briefer pe…

"let me tackle the CAPTCHA" extremely 2026 energy

20 likes 1 RT

Replying to M. Friedenbach 🏴‍☠️ @MarkFriedenbach ·

@nosilverv @aphercotropist @glorpalorb @QiaochuYuan The zizians experimented with polyphasic sleep. Yes, the Bay Area rat group that went on a murder spree. Not in 2013, but a little bit after.

@MarkFriedenbach @nosilverv @aphercotropist @glorpalorb @QiaochuYuan they experimented with much much weirder things, such as trying to sleep only half your brain at once (plus terrible philosophy)

4 likes 0 RTs

Replying to QC @QiaochuYuan ·

hey btw don't actually switch to polyphasic sleep. a group in the bay area rat scene did a long experiment with this in 2013 and several of them went crazy. especially in the context of, y'know, *sweeps hand around* you are just begging to induce actual psychosis in yourself

quoting @sama: "post-AGI, no one is going to work and the economy is going to collapse" "i am switching to polyphasic sleep because GPT-5.5 in codex is so good that i can't afford to be sleeping for such long stretches and miss out on working"

@QiaochuYuan I tried extreme polyphasic multiple times, it taught my body how to nap, which let me sleep biphasically for years (~6h/night + 20min midday nap) and actually fall asleep whereas when trying to sleep 8h/night I had trouble falling asleep at the same time (delayed phase? idk)

4 likes 0 RTs

Retweeted @TylerAlterman

What about instead of a watered-down overgeneral Unitarian Universalism, you have a congregation of smaller groups that…

Retweeted @TylerAlterman

We live in an ever-changing stream that asks for our continual evolution. We can take great inspiration from the past bu…

@QiaochuYuan but to be clear I agree that combining not-much-sleep with constantly-talking-to-LLMs does sound like a recipe for psychosis

2 likes 0 RTs

Replying to ꜱᴘᴀᴄᴇ ᴘᴜɴᴋ @_space_punk_ ·

I was just thinking this morning how the breakfast question is another example of the IQ bellcurve meme You actually kind of can't change single variables. To make an apple pie, one must first invent the entire universe

quoting @ElonBachman: Newcomb's Paradox is the Breakfast Question for rationalists https://t.co/mDMwHVscyf

@_space_punk_ this is a great point that is not widely understood. you might also like these tweets about Newcomb's problem

Jakeup @myhandle ·

as Emmett notes: if you believe that you are *somebody*, i.e. a type of person, someone whose decisions and behaviors are somewhat consistent (and thus predictable), one-boxing is easy but people two-box because usually thought experiments you are NOBODY x.com/eshear/status/…

quoting @eshear: @St_Rev It simply points out that existing decision theory does not take into account the idea that cognition is embodied; that the kind of thing you *are* can be as important as what you *do*.

1 like 0 RTs

Replying to Tyler is finishing a book, slow to reply @TylerAlterman ·

Which individuals had the absolute biggest positive impact in history? Give me your top 3

@TylerAlterman butterfly effect makes this very very weird to think about. once your effects get large enough, they're upstream of bad things that happened too. suppose hitler would have died in childhood if not for a vaccine. not saying impossible but weird

ꜱᴘᴀᴄᴇ ᴘᴜɴᴋ @_space_punk_ ·

I was just thinking this morning how the breakfast question is another example of the IQ bellcurve meme You actually kind of can't change single variables. To make an apple pie, one must first invent the entire universe

quoting @ElonBachman: Newcomb's Paradox is the Breakfast Question for rationalists https://t.co/mDMwHVscyf

4 likes 0 RTs

@TylerAlterman the weird dimension I actually wanted to get at was something about the "individual" thing. eg I'd be tempted to say St Paul over Jesus, because without Paul the church would not have developed or spread so well, and might have just kind of collapsed claude.ai/share/830033df…

2 likes 0 RTs

Replying to Guy @nosilverv ·

@AnnaLeptikon Very nice prose. But why after death? Unless he means to be twice-born. "The Kingdom of God is Within You"—it's here, right here, not far away, in the most intimate place you can look. Samsara is Nirvana.

@nosilverv @AnnaLeptikon the whole point with making the analogy to ducks is that water fucking EXISTS ON EARTH "a man has the desire to suck his own dick. This isn't possible on earth for most men, so God created a parallel universe where valve-style portals, so man can suck his own dick" = same logic

1 like 0 RTs

@nosilverv @AnnaLeptikon Like really this whole line of reasoning is a knockdown argument AGAINST an afterlife or some transcendent otherworldly solution although indeed it does suggest that whatever boring immanent situation you have so far isn't cutting it and that something else is possible on earth

1 like 0 RTs

@tasshinfogleman @dthorson (to be clear, I don't usually put people's essays into gptzero, but it felt very AI-pattern words, so I was curious what the % was and was actually kinda surprised to discover it came out at 100%, not just substantial edits)

0 likes 0 RTs

Replying to David Shapiro (L/0) @DaveShapi ·

@VahnAeris @Liv_Boeree Huh I thought it was @jim_rutt but I looked it up and you're right

@DaveShapi @VahnAeris @Liv_Boeree @jim_rutt they co-developed the idea of Game B I believe, along with Jordan Hall

1 like 0 RTs

hmm these models need more RLHF (Reinforcement Learning from Having Fun)

7 likes 0 RTs

talking with lawyers more recently as part of my job (corporate stuff) and one useful concept I've gained from that is 'fact pattern' - before interpretation/judgment/ruling, what is the not-in-dispute situation? phrases like 'I'm familiar with this fact pattern'

10 likes 2 RTs