Matt

@DaveShapi With the assumption that this premise is vakid and the economics don’t check out, what are you expecting people to do when their labor is devalued? What are your thoughts on a permanent underclass?

I asked GPT pro if it was remotely possible for the attention/meaning/experience economy to provide enough household income, assuming that AI and robots ultimately do the vast majority of agriculture, manufacturing, and services jobs. The answer is amazing. ----- Short answer No. Even under generous assumptions about how much time people devote to human‑provided experiences, a “meaning/attention/experience” labor market cannot, on its own, generate a wage bill large enough to support most households. The binding constraints are arithmetic: finite human attention, strong scalability of the most popular experiences, and winner‑take‑most revenue distributions. When you quantify those constraints, the number of jobs and the total labor income that can be sustained by human‑delivered meaning tops out far below today’s household wage mass. The binding constraints in plain economics The first constraint is the 24‑hour day. In a fully automated economy, suppose the average adult devotes a very large 20 hours per week to paid human‑delivered experiences (live events, human‑led classes, coaching, small‑group dining, in‑person arts). With roughly 260 million U.S. adults, that yields about 270 billion consumer‑hours per year directed at human performers or hosts. To turn those hours into household income you need two things to be simultaneously true: a high price per consumer‑hour and a high labor share of that price. Let be the annual wage bill you want meaning jobs to finance, the average price paid per consumer‑hour for human experiences, and the labor share of that revenue. Then W \;=\; \lambda \cdot p \cdot \big(N \cdot L\big), If you ask this sector to deliver even a modest wage bill of $4 trillion per year—roughly a fraction of today’s aggregate labor compensation—then with billion hours and a generous , the required average price is about $24.65 per consumer‑hour. At $8 trillion it rises to $49.31 per hour, and at $12 trillion to $73.96 per hour. Those prices would have to be paid by everyone, week after week, for a very large portion of their leisure—despite the existence of near‑zero‑cost, very high‑quality machine experiences as substitutes. Raising to 40 hours per week helps only linearly; even then you still need per‑hour prices most households will not pay at scale. The second constraint is labor intensity versus scalability. Jobs come from experiences that are not scalable. If one human can simultaneously entertain 1,000 people online, each performer‑hour supplies 1,000 consumer‑hours and absorbs very little labor. To maximize jobs you must push the mix toward low‑audience, in‑person, bespoke encounters—1:1 lessons, small tables at restaurants, intimate theater. You can formalize this with an audience‑per‑worker ratio . The number of jobs sustained is \text{Jobs} \;=\; \frac{N \cdot L \cdot f_H}{A \cdot S}, The third constraint is superstar economics. Attention markets are heavy‑tailed. A small fraction of creators captures a large fraction of revenue, leaving median earnings thin. Your own empirical ratio—30–50,000 hours of captured attention monetizing to $2,000–$2,500, or roughly $0.04–$0.08 per attention‑hour—is entirely consistent with that distribution. Aggregating such low ARPU across national time budgets does not produce enough labor income unless prices per consumer‑hour are forced dramatically higher and spread widely across the population, which runs directly into the substitution effect created by cheap, compelling machine experiences. Why “more leisure” does not solve it It is tempting to say that if people stop working they will buy far more experiences, and therefore the sector will scale. More leisure does expand quantities demanded, but it does not loosen the core constraints. If the experiences people most enjoy are scalable (streaming, esports, mass festivals) the audience‑per‑worker ratio rises and employment intensity falls. If the experiences are kept intimate to maximize jobs, the required spend per hour must rise to levels that, in the presence of excellent machine substitutes, most consumers will not pay. Either way, the product of “time × willingness‑to‑pay × labor share” does not reach the trillions needed to replace the wage mass of the rest of the economy. What would have to be true for a “yes” A “yes” only emerges under conditions that contradict the scenario’s own economic logic. You would need most consumers to devote a very large share of their leisure to low‑scalability human encounters, to pay high prices per hour for those encounters, and for a very high share of that spending to flow as wages rather than to venues, platforms, IP owners, or capital. You would also need attention markets to not be winner‑take‑most, so that earnings are broadly distributed across millions of workers. In a world where machine experiences are better, faster, cheaper, and safer, none of those conditions holds without external compulsion or subsidy. Bottom line On its own terms, a meaning/attention/experience labor market cannot shoulder the primary role in household income. The arithmetic of finite attention, the economics of scalability, and the distribution of attention‑driven earnings together put a hard ceiling on both the number of jobs and the total wage bill that this quadrant can support. Your intuition that such work could plausibly employ on the order of 10–20 percent of people, with the remainder supported by other mechanisms, is consistent with the quantitative bounds.

@banteg Its more like a large body of water flowing down hill. It doesn’t have a theory of mind, but its a lot of power moving towards some optimum

explain why in all ai-kill-everyoneism scenarios the ai possesses god-tier engineering skills while at the same time being a mindless objective-driven drone with zero self-awareness. you know what else is underpinned by never questioning the premise? fiction! if anything, this is the type of scenario we can get into from ai safety people hardcoding their wet fantasies into the machine.

@4be1 @bryan_johnson Google russell conjugation

@bryan_johnson What's inaccurate about the current title?

Applications to work with me have increased.

@max_spero_ @notnotstorm I think you’ll find the real arguments lie in your last sentence here. It >was< sufficient, but its becoming increasingly less so. “Most” is very hand-wavy. Who is an authority? I personally don’t trust the US government for anything beyond their monopoly on violence.

@notnotstorm I understand that, but there are near zero real world use cases for distributed trustless consensus. Most of the time an authoritative single source of truth is sufficient

I was a TA for Stanford's 2018 class on cryptocurrency. I helped teach the math of blockchain, cryptographic primitives that make up Bitcoin, smart contracts etc. But every office hours I would field questions from struggling students asking which coin they should buy. I kept telling them it's not an investment vehicle and I don't hold any crypto, but it didn't really matter. Bitcoin kept going up but I never was able to come up with a real world use case where blockchain was better than a database. It's unfortunate that the technology is so elegant but so limited in its uses. Condolences to those in crypto now but this has been the writing on the wall for a long time. It's all speculative pump-and-dumps and you haven't been honest with yourself if this is the first time you're recognizing it.

@War4theWest @sethmr @BlakeTOliver @the_adambrooks The argument of “we did it for decades” is not good either. We used to live in the wild and hunt, why go into an office work on abstract things at all?

@sethmr @BlakeTOliver @the_adambrooks Not in the least, but wow, are you triggered by the idea of actually having to show up for work physically. Maybe Elon was right and we should just replace all of you...

JPMorgan just told all 300,000 employees they must return to office full-time in March. Their response to employees raising concerns about childcare and commuting costs? They shut down comments.

@RonanKMcGovern @DavidDeutschOxf But let’s think like an engineer instead of a philosopher - if a system thrashes between poor rulers then that system is inherently bad, regardless of ruler. The voting system is more important - simple majorities and first past the post voting schemes are a problem.

I came across this 1988 piece in the Economist by Karl Popper – via David Deutsch below. I wrote about Proportional Representation very recently, so I'll avoid repeating what I've already said. The old question. The old question is "who should rule"? Should it be capital or labour? Should it be the rich or the poor? Should it be the wise or the many? Even if it should be the competent, the wise and the virtuous that rule, who ultimately decides what is virtuous? And, what is to stop virtuous rulers from becoming self serving and evil? The new question. If we don't know or can't agree who is virtuous, perhaps we can focus on a more practical question instead – "What is a system that allows an incompetent or evil government to be replaced?" This, is the new question. This is the reason why Popper supports democracy – because it allows a government to be overthrown without violence – to be replaced by a simple majority, a vote of over 50%. "My theory of democracy is very simple and easy for everybody to understand. But its fundamental problem is so different from the age-old theory of democracy which everybody takes for granted that it seems that this difference has not been grasped, just because of the simplicity of the theory." – Popper "All these theoretical difficulties are avoided if one abandons the question 'Who should rule?' and replaces it by the new and practical problem: how can we best avoid situations in which a bad ruler causes too much harm?" – Popper Why Proportional Representation does not distribute power proportional to votes, and worse, makes it harder for governments to be replaced. Proportional representation stems from the belief that the power of each party (or candidate, in single transferrable vote systems) should be proportional to the number of votes each receives. Yet, this is not what proportional representation achieves. Instead, it encourages coalition governments that form or break apart based on small parties with disproportionate power. Worse, it prevents bad governments from being replaced. If the virtue of a system of governance is that power should be proportional, a two party system is more proportional in how the power of government relates to the votes of the people. But, this misses Popper's point. Popper's emphasis it is that democracy – and a two party system in particular – is virtuous because it is most effective in allowing bad governments to be replaced without violence. ctd…

@pee_zombie One over many

wait monopoly is kind of a weird word. mono poly. what's up w that

@levelsio Public reviews that cant be corrupted by a monetary incentive, or gamed by the entity under review.

Why hotel photos always lying And how to fix this?

@normal_brandon1 @Aella_Girl Probably something related to law of large numbers / large distributions

@Aella_Girl Why does this work so well?

Imagine a circle, with a little spinner on it that you can flick. One quarter of the circle is yellow (25% of it, like a big pie slice) and the rest of it, three quarters (75%) is red. You flick the spinner. It spins, and slowwwly stops. Where did it land?

@genuinearticles @bryan_johnson Pareto is an observed power law distribution

@bryan_johnson What's the difference between the Pareto principle and power laws? Or is one just an application of the other?

This is my read of the health landscape and how we win. We’ve been going through a period of health awareness and education for the past few years. The power law principles of health are known. We know what 20% of healthy behaviors delivers 80% of the benefits. Learning about health is easy; doing those things is hard. Doing is also the most valuable. In fact, you don’t even need to learn in order to do. Doing is value. We will succeed at doing by building community and societal systems that help people consistently do the right things for their health. Giving them an edge to overcome bad habits and create new good habits. These communities need to be large and strong enough to create in-group defiance towards outside groups who seek to undermine the new emergent culture. They'll need a strong belief that they are correct in their health choices and martyrdom culture is wrong. Prioritizing health is much easier when it’s done with others. Health is particularly hard for so many because of individual isolation, how easy it is to do harmful things to oneself and loved ones, and the risk of being ostracized by social groups for not participating in Die culture. These health communities need to have rewards, status, achievement, and growth opportunities. They need to invite pro-social behaviors where members care for each other and invest in the community. Eschewing the cynicism that primarily drives online interactions. This is what my team and I are building right now. Expect to join soon. Enough of us will gather together to create strength in numbers. We will build the future of human together. Others will join in time too because sometimes we embrace the future and other times the future pulls us kicking and screaming into their magnificence.

@evan_van_ness It must have been a reddit or twitter thread, but the suggestion that the olympics should really split into 1) pure human, completely raw dog training, and 2) accelerationist style push-the-limit enhanced human cyborg experiment-at-your-own-risk. Sounded cool to me.

Trying to figure out why 1. Cold War is over. Olympics felt a lot more meaningful when it was a proxy event for capitalism vs communism 2. No longer even an amateur event, half the sports are just flat professionals 3. Childhood belief that steroids are not widespread. As I got older I realize many events are just "who can dope and beat the drug tests" competitions as much as athletic 4. Corruption by Olympic committees. Obviously there's billions of dollars in play and people want to get paid, but the "it's all so pure" narrative they sold me hard as a kid was nonsense

The Olympics feels extremely Boomer to me

@Sithu_Aye Synth VST recs? Serum/Vital Wavetable tutorials you like, or any tips that made building synth layers easier / faster?

Songs, like onions or ogres, have layers. These are the layers of the chorus of my song 'Obsidian' from my new EP 'Kindness'

@GrantSlatton @gfodor This triggered a line of thinking for me. Some combination of negativity bias + signal dampening is happened on social media with downvotes and I think thay reduces the overall SNR. If twitter can figure out the bot problem I would drop everything else I think.

@gfodor I got a new laptop a year or two ago and just never bothered to log back in Reddit and HN are just so inferior to (properly used) twitter in almost every way I think twitter's lack of downvotes is actually a huge part of that

Deleted my HN account - it’s over

@algekalipso This is very interesting. Would be worthy of a 3blue1brown style node/edge graphic - at least thats how I was keeping track of the processes described.

You've heard of Marr's 3 levels of analysis? Inputs-outputs, Algorithm, Implementation ------- Let's say 3 aren't enough. Alright, here's Emilsson’s 17 levels of analysis: The brute ontology level The awareness and attention over the objects of the brute ontology The awareness of this awareness The transmission of both the original awareness and this meta-awareness to a set of sensory modalities Turning the cluster of attention and awareness in the sensory modality into the ontology one is expecting via sharing the same interlocking symmetries (this will turn into inductive priors at a functional level) Awareness over the gestalt of this process Sequencing of such gestalts into ordered events Overlapping ordered events creating tree structures creating “counterfactual space” Awareness over this counter-factual space as an object of attention Mapping of self-aware counter-factual (phenomenal) space into a causal hypergraph Causal hypergraph used as model to capture empirical data (which is also defined in terms of its own ontology along with all that entails) Cross-validation of model and empirical data (this itself requires a whole datastracture to become self-aware, complicating matters a bit) Awareness over the attention over cross-validation comparisons (which will depend on an ontology of meta-counterfactuals) Awareness over attention focused on awareness simultaneously with the awareness over the cross-validation Symmetry measure change due to the simultaneous awareness on this combo object (the previous line) in terms of total symmetries erased, gained, or broken Awareness over multiple of the objects described above Awareness over the change in symmetry features as a function of comparisons between objects of the previous line ----- I thinks this is what we do when we analyze an information processing system in the hopes of deciding if it's conscious or not.

@julesterpak @kylascan Did you just finish the book or did a particular insight trigger this tweet

Amazing @kylascan

@julesterpak Seems like a good take. Not just cross platform, but across echo chambers and filter bubbles. If my parents, girlfriend, friend groups, all see the same meme/news/event as me I would consider thst viral. Might have to start categorizing viral types.

Hot take but it’s almost like even though MrBeast videos gets tens or hundreds of millions of views, the only ones I consider viral are the ones that spark cross-platform commentary that breaks the video out of that new YouTube user recommended page type of audience bubble

People used to deem something “viral” after it hit a large view or like count, but with the rise of TikTok and how often people now hit those thresholds, the best way to deem something “viral” relates back to the word’s roots It’s about the cross-platform appeal of a piece of media and the commentary it spurs via stitches/duets, quote tweets, YouTube commentator coverage, articles, and/or so on. If the media just has a large view or like count, it’s likely just in a niche bubble

@santi_monse @craiglatzke @S_R_Angelov @0xgaut skill issue

@craiglatzke @S_R_Angelov @0xgaut A tool that BURNS the coffee! Any other method is better!

Remember, if you want to save money, just make your coffee at home The coffee at home:

@KlompWerner @fori920 @chaseedso @zacbowden @remixedcat New platform supports old apps via native binaries or emulation layer. Directionality is a relative.

@fori920 @chaseedso @zacbowden @remixedcat Going to arm is backwards?

All of these apps are guaranteed to work on Windows on Arm.

@algekalipso I’m in a punk band open.spotify.com/artist/2pywJgB…

Do you make music? Please leave a link in the comments for other qualia kids to check out 💗