phrog

[Chris Freiman @cafreiman]

Capitalism doesn’t *require* endless growth, but Paul Romer explains how (effectively) infinite growth is possible:

[🕊️ @lichthauch]

Someone told me years ago that you cant think your way into faith, that your head has to shut up for your insides to hear anything. and i said thats stupid, thats irresponsible, you dont turn off your brain to approach truth, you use it, thats the whole point. he just nodded. didn't argue. i didnt know then that he was nodding because he recognized me. took me way too long to get that. years later, after losing enough, i was reading something i'd read twenty times before and for once i wasnt analyzing it, i was just holding it, i didnt have the strength to do anything else honestly. and it grew. same words as always except they didnt stay on the page this time i dont really know how to explain that part. anyway. now people argue with me the same way. that wont work, thats irresponsible, you dont turn off your brain. and i just nod

[🕊️ @lichthauch]

The specimens of men who imagine they can desert their purpose by wrecking their bodies are the most pitiable creatures in existence there is nothing more grotesque than a gifted man with glazed eyes and soft hands rotting on a couch in his rented quarters. purpose is a hereditary deformity of the soul that you were built to perfect and the man who mutilates himself to escape it inherits the worst version of the same condition a burden with nowhere to sit except directly on the flesh of his disgrace

[JT @jiratickets]

this might be the best tweet ever written

[Just Another Pod Guy @TMTLongShort]

Instead of focusing on the whole “Elon is a genius vs Elon overpromises” debate you should instead ask yourself “what if the rapid solving of mathematics is going to quickly be followed by the wholesale solving of the field of physics” and in turn what that means for your assumptions around ASMLs moat.

[yung macro 宏观年少传奇 @apralky]

This is such an interesting ever-present line of rhetoric now – seemingly these ppl don’t realize that they’re not discovering some novel pathology incubated with Elon’s Twitter takeover in 2022 – and that an increasingly enfranchised “market of ideas” progressively skewing general consensus toward slop has been a front page issue since like the 20th-century Marxists held a say (and certainly well before then if you count Enlightenment Liberalism as slop as did many of the contemporaries).

I’m not sure to what extent the team at X is heavy-handed with the algo, but it isn’t difficult to understand that an organically democratic sorting process will favor dog pictures and Kalshi clickbait over NYT articles, just as it favored NYT articles over French sociologists or whatever as an earlier stage of the exact same drift.

The corollary is obviously that not only is the issue not new, but neither is the proposed remedy – that of some central authority inorganically suppressing the demotic aspects of the discourse and/or inflating favored ones. We call this censorship, we’ve tried it to varying degrees, in the limit we agree that it's a very bad thing. Moving up the centralized curation spectrum comes with the exact same tradeoff as in the limit cases, with an appropriately scaled magnitude. There is a very clear marginal cost associated with every marginal unit of the benefit. Arguing that more censorship is optimal because it brings the benefit with no mention of the cost is clearly not the right way to model this.

The more naive elements perhaps think no such organic drift exists in the first place and the progressive degradation is a result of this or that contingency, if Elon takes his thumb off the algo THEN it will all go back to normal and the 26-year-old plumber from Kansas will put down the conspiracy tiktoks and pick up a Chad Jones PDF, but this angle is straightforwardly not serious. If the plumber no longer opting for Ian Carroll is our target variable our policy set is firmly limited to inorganically suppressing his access, in whatever format we might opt to do that.

There was a funny episode of this where Douglas Murray with his usual suffocating condescension lectured Joe Rogan on the show about how he should be more careful with the sorts of characters he “platforms” – why not have more experts on? And so on. Joe couldn’t really muster a coherent response, but isn’t Douglas also confused here? We have the expert platforms, and people picked Joe Rogan. The expert platforms were an option. Nobody forced Joe Rogan onto the people. They’ll keep picking Joe Rogan until we take away Joe Rogan. If Joe Rogan in a moment of moralized enlightenment decides to pivot to running an expert platform – they’ll start picking Shmoe Rogan who hasn’t. This will naturally keep happening unless we're willing to suppress the market’s freedom to satisfy that demand. We haven’t agreed to do that, and we probably wouldn’t if someone were to pose us with that specific policy set. We'd probably call them a fascist, maybe rightly so! Instead we can keep our heads in the sand firmly convinced that the toddlers’ choice of candy for dinner is a product of artificial intervention – complaining about obesity while refusing to take candy off the menu. This is unproductive.

[yung macro 宏观年少传奇 @apralky]

Good & interesting piece, but a necessary caveat is that it’s essentially a hypothetical conditioned on severe institutional failure, rather than some sort of macro inevitability.

As the piece itself notes, one of its two failure modes -- liquidity stress and capital impairment -- includes a liquidity component the Fed can address quickly with liquidity facilities and asset purchases (repo lines, QE...), as seen in recent episodes of banking stress.
Losses from impaired assets won’t disappear, but can in principle be moved where the broadest shoulders are, of which there will be plenty in this hypothetical economy (much more than before! Couldn't they just tax Anthropic into paying all your pensions?). This would be done through regulation or a fiscal mechanism.

The second failure mode -- an aggregate-demand shortfall from mass unemployment -- can be addressed through fiscal policy (transfers, wage subsidies, etc.). The piece argues this may be constrained by falling tax revenues, but in a deflationary/low-inflation environment the Treasury can run large deficits, and the Fed can buy ~as much of that debt as it needs to, leaving aside that we’d expect the tax structure to change. The political process is unlikely to be as meaningful a bottleneck as the piece claims, as the hypothetical fiscal hawks pointing to unsustainable deficits would not have much of a point, given the economy in this hypothetical will have a much higher potential output (also, that obviously wouldn't poll well).

The first principles intuition obviously tells you something is awry when we’re told that people will want (at least) as many real things as before, and the economy will have the means to produce more real things than before, but the people won’t be getting the amount of real things they want or need. This is because that typically requires major policy/institutional frictions (or delay) in translating capacity into purchasing power.

If you suddenly have 2 loaves of bread instead of 1 in your house, and you weren’t starving with 1, you probably shouldn’t starve when there’s 2. But if -- by some hypothetical -- through the complexity of the novel 2-breadloaf production process you suddenly get tangled and can no longer access the cupboard, then it’s quite possible you will starve.

[phrog @qntfn]

@GWHayduke97 @WillManidis Tell me you understand nothing about the problem without telling me

[roon @tszzl]

the acceleration is perhaps the last most aesthetically interesting thing happening on earth and is slightly vampiric in that it makes everything else slightly less aesthetically interesting

[phrog @qntfn]

@parasolstand @jeremykranz @DavidSacks @SemiAnalysis_ *** 2024, rather

[phrog @qntfn]

@parasolstand @jeremykranz @DavidSacks @SemiAnalysis_ Interesting. So pointless that it won the 2023 Nobel Prize in Chemistry?

https://t.co/RMIlVm9Jhp

[yung macro 宏观年少传奇 @apralky]

moving past the grating kitsch of these geriatric tantrums for a moment, it will never not be hilarious the extent to which the sub-intellectual populist left's rhetoric has always been a product of their complete incapacity to intuit sustained economic growth at any level of abstraction whatsoever or to generalize beyond like one node of causality (their more rigorous "theorists" to their credit didn't seem to struggle with this)

they mean well! they just aren't very smart

notice how the push-back isn't that the technology is pointless and a waste of time, but rather stems from an admission that it would be unprecedentedly transformative and disruptive -- and this is precisely why it's best to maximally keep it in check, in order to preserve some ill-defined sentimentalized notion of the presently prevailing "social fabric"

"remain human" at the cost of what you clowns? what exactly is it you're "rebelling" against? 5% GDP growth? a collapse of extreme poverty from near-universal to under 10% of the global population? vaccine distribution halving childhood mortality in Sub-Saharan Africa? antiretroviral drugs cheap enough that HIV is not certain death anymore? agricultural yield growth? food for the global poor? tens of millions spared from water scarcity? families no longer needing their children working at subsistence margins? refrigerated insulin and blood supplies? mothers actually living through childbirth? and for what exactly? so you can keep your SAG-AFTRA job? easier book deals? professional status preservation for a few thousand millennials in New York? some provincial fantasy of giggling couples in tulip farms? a 70-year lifespan? is 5% infant mortality in Afghanistan a part of "being human" too? genuinely what is this "humanity" that's perceived to be at stake here?

[tomie @tomieinlove]

One problem with the argument that “AI can’t generalize” is that humans also can’t generalize. We mostly just train on the test set.

This isn’t just conjecture, mind you. It’s proven. There’s a famous philosophical problem, Molyneux’s problem. It asks if blind people, who identify 3D shapes like cubes or cylinders by touch, would be able to, if cured of their blindness, recognize those shapes by sight alone.

The answer is no. In the early 2000s, we cured some people blind from birth and gave them that exact test: identifying 3D shapes by sight which they’d previously felt. The subjects were helpless to connect the cube they’d once held with the cube they were seeing, helpless to differentiate it from even a sphere.

This explains why it’s so rare to see humans, even our best and brightest, making genuinely deep connections. We can’t. We don’t have the architecture, the power. We’re remarkably primitive.

[phrog @qntfn]

@semiDL @tszzl cool graphic. may i ask where it’s from

[roon @tszzl]

Til the rains do fall on Olympus Mons, and our descendants can chud out about offworlders arriving only after the hard work of terraformation is done

[roon @tszzl]

the primary criticism of AI you hear has nothing to do with water use or existential risk whatsoever: most people just think it’s fake and doesn’t work and is a tremendous bubble eating intellectual property while emitting useless slop along the way. when GPT-5 came out and perhaps didn’t live up to what people were expecting for a full version bump, the timeline reaction was not mild, it was a full-scale meltdown. there are many intelligent (and unintelligent) people who latched onto this moment to declare AI scaling over, thousands of viral tweets, still a prevailing view in many circles. The financial-cultural phenomenon of machine intelligence is one of the most powerful in decades, and there are a lot of people who would like for its position to be weakened, many outright celebrating its losses and setback. Michael burry of ‘Big Short’ fame, unfortunately the type of guy to predict 12 of the last 3 recessions, has bet himself into insolvency on the AI bubble’s collapse

one of the stranger things about this time is that there are very few secrets, and very little reason to be so misinformed. model labs have very little space in between creating new capabilities and launching them to the public. The view among the well informed public and not just “lab insiders” is that machine intelligence is absurdly joyfully smart at so many new things every month. It’s actively contributing on the cutting edge of programming and math and science. Sebastian Bubeck and co’s recent paper reports that GPT5-pro is capable of producing results on the frontier of theoretical physics research, Terry Tao wrote a blog about “vibe-proving” Erdos problems with the auto-formalization AI Aristotle. You can read that these scientists are using it to actively contribute to black hole physics, tighten mathematical bounds in optimization theory, churning morasses of biomedical data into real insight. Google Deepmind, from the way they are signalling, seems to be slowly closing a dragnet around the Navier-Stokes smoothness millennium problem (though of course, I don’t know). Several companies stocked top to bottom with brilliant scientists are racing to build pipelines to solve novel physics and chemistry and biology

You can read online about the new kinds of organizations being born around machine intelligence as a first class factor of production. For the first time, the new factor actually gives you ideas for improving the processes themselves. It’s designing whole assembly lines where some of the workers on the assembly line are also AIs, and the line itself is morphing and self-optimizing. Tiny teams are producing amounts of work that seemed impossible to organizations of a few years ago. It’s hard not to feel excited by the productivity growth happening in these admittedly narrow software sectors. Every time I use codex to solve some issue late at night or GPT helps me figure out a difficult strategic problem I feel: what a relief. There are so few minds on Earth that are both intelligent and persistent enough about some intellectual pursuit to generate new insights and keep the torch of scientific civilization alive. Now you have potentially infinite minds to throw at infinite potential problems. Your computer friend that never takes the day off, never gets bored, never checks out and stops trying. You can feel the unburdening of Atlas, the takeoff. It feels more prosaic and less poetic than it did in 2023, even though the results speak for themselves more loudly

[Michael Ferguson @TheRealFergs]

Loving that Disney keeps dropping these kind of papers: https://t.co/1NQHRiuGdZ

[Just Another Pod Guy @TMTLongShort]

Whenever I feel anxious that my views on AI are misaligned with consensus and I may be missing something I remind myself that it took the Street nine months to take the time to read Situational Awareness, weeks to appreciate Deepseek, and interviews with Lukasz at OpenAI or Jeff Dean at Google routinely clock sub-10k views on YouTube even weeks after release.

A large part of the market flows is being directed by boomer PMs who don’t know shit and midwit analysts who think they’ve done sufficient diligence by listening to a Chanos podcast, skimming a badly distorted summary of what Ilya said, and reading a sellside primer or two.

Think for yourself - dispersion increases from here and we are so early.

[near @nearcyan]

you should be writing quarterly shareholder updates even if you don’t have any shareholders. even if the only thing you’re working on is yourself