← Influencer libraryEdition 2026-10-10, preview.
THE LIBRARY

Balaji Srinivasan

1,539 items collected

1,539 matching items · newest first

XThread

Before communism, Russia was good at capitalism. https://t.co/DkQCqCmVWN Few truly understand what communism was. Previously functional civilizations were wrecked by what amounted to a fundamentalist takeover.

XThread

Don’t be surprised if China offers Terry Tao the chair of the mathematics department at Tsinghua University. And all the funding he wants. Because it is possible for the US to fumble the bag as the destination for global talent. And indeed it is fumbling that bag right now. Remember, China is focused on recruiting global talent now. They recruited another Fields Medalist from Harvard, the Yau of the Calabi-Yau manifold. https://t.co/7V73KcdYhs https://t.co/tAeK92L0rY In a recent shift, China has also begun seriously recruiting tech talent of non-Chinese ancestry. Their new K-visa is focused on this. This is another pickup for Tsinghua: …

XThread

This is on track to happen. And when it does, all the anti-tech radicals across the political spectrum will regret all the anti-Elon stuff. But too late. It’s extraordinarily difficult to build something like SpaceX even in peacetime. If huge swaths of society are irrationally trying to get you, the difficulty rises further by orders of magnitude. Meanwhile, China just plows ahead relentlessly.

XThread

On many graphs of the physical world, China is in first place by a wide margin. But if you look more closely, India is a distant but real runner up. https://t.co/0jy25Cvb8R The same pattern holds in nuclear. China is #1 in reactors under construction. But India is in second place. https://t.co/dVuLKe1TZW Electricity generation is similar. China’s recent increase is unparalleled in history. But India is increasing quickly, and will flip the EU soon. https://t.co/yRWP6UzBgX In smartphones, China is again global number one. But over the last few years, India has risen to become the #2 manufacturer. More than 20% of iPhones are now made in I…

XThread

Many 20th century systems will melt down in a low-trust, high-tech environment. Fraud rates will just get out of control. We’ll need to rebuild those systems with cryptography. Digital signatures are much harder to fake than signatures. In the medium run, states will need to replace the signature-checking step with a digital-signature-checking step. In practice, that could be as simple as having Texas state incorporate Google/Apple-style auth, and then require digital actions on deeds to be authenticated via passkeys and/or Yubikeys. This can be generalized. The state becomes a digital platform and any updates to official registries requi…

XThread

We’ve seen this movie before. After they defeated the far left, and wound down their global empire, Russians returned to putting faith in almighty God over their failed State. https://t.co/YK3WyyLZRj Remember, communism was the god that failed. Just like wokeness, the far left tried replacing faith in G-o-d with faith in G-o-v. Yes, it’s true that communism is centralized left and wokeness is decentralized left, but for the purposes of this discussion those are details. Both doctrines elevated the State over God as the central force that should always do something. For example, think about how a woke mocks “thoughts and prayers” after a t…

XThread

The war for talent is on. China already took over US manufacturing. Now it's making a bid for the talent pipeline too. https://t.co/dFrlMSJOwp There was a bad argument that went something like this: "if it was smart to recruit great engineers from around the world, why isn't China doing it?" It was a bad argument on three levels. (1) First, China actually did send its own engineers around the world to learn, for decades, just like the Japanese did decades before them. They didn't insist on 100% homegrown talent in every respect. (2) Second, until recently, there was widespread denial that China had even succeeded technologically in any re…

XPost

@NSmolenski A few points. TLDR: my post advocates for understanding tradeoffs, not empire per se. (1) First, please reread the post. It's not "advocating" for empire. It is, at most, devil's advocating for empire. That is, it's pointing out that one should consider that (a) the American empire exists, and (b) generates enormous amounts of money for at least some Americans via money printing, before (c) sunsetting it. (2) As context, most people either deny that empire even exists, or only point out the costs and not the benefits. Now, it is certainly possible that sunsetting the empire is still desirable on balance, just as sunsetting the …

XPost

What really happened with Windsurf, Figma, and the FTC? New pod with @stevesi & @eriktorenberg of a16z. 2:25 Techlash 3:23 No regulators for computers 6:04 Network vs state 25:02 M&A review 32:04 M&A power law 44:05 What happened with Windsurf? 01:14:16 Winning by legislating https://t.co/vP5mSbmXy0

XThread

Here's the podcast that @BritishHodl and many Bitcoiners liked. 0:00 - Introduction 1:32 - West vs East 6:33 - The ascending world 10:30 - Denying while declining 14:00 - Millionaire migration map 18:20 - Several exponentials 20:10 - Four factions (Blue, Red, China, Internet) 25:00 - Red America v China 33:05 - Blue America v Internet 39:35 - Geographical splits 43:52 - Dollar inflation is global taxation 49:30 - Left and right ending their empire 1:07:20 - Superbitcoinization Is already here 1:26:00 - The Internet frontier 1:32:00 - Sovereign debt crisis 1:42:00 - Full faith and credit 1:45:00 - Where's talent moving? 1:48:00 - On India 1:5…

XThread

In general, AI text should not be used verbatim. In writing or in speech. Think of it as lorem ipsum. Or more precisely, lorem aipsum. It's placeholder text. All this may change as tech evolves. Perhaps everyone gets their own highly personalized writer, such that the AI copy they emit isn't cliche-ridden and overwritten. But at least for now...there's an instant turnoff when you see someone use AI text verbatim in an application form, or a Substack post, or a tweet. It just signals that they can't write well, or that they're lazy, or something. To be clear, AI is useful for coding, for data analysis, for snippets of legal work, for situat…

XThread

The Soviets went nuts on money printing in the years before the collapse. https://t.co/ccDFEykUlY From @ryanmcmaken in 2022: "It took some time for the official numbers to catch up with reality. The regime’s official numbers had long understated even the moderate levels of price inflation in earlier periods, but after the mid-80s, the gap between official inflation and estimated real inflation grew considerably. Efremov summarizes the divergence, noting that in 1988 official inflation was 0.6 percent but 6 percent in the real marketplace. By 1989, official inflation was 2 percent, but it was really 8 percent. In 1990, it was 5.3 percent, b…

XThread

Ants are dumb...but ant colonies are smart. More generally, there's the idea of swarm intelligence. Quite a lot of research in this area. Might be applied to crypto and AI agents specifically. And distributed systems more generally. The swarm intelligence algorithms have amazing names. They're all named after bees, fireflies, wolves, whales, dragonflies, cuckoos, and the like. It does feel there's something here in terms of getting crypto crowds to cooperate, and getting AI agents on the same page. CITATIONS (1) Particle Swarm Optimization (PSO) Kennedy, J., & Eberhart, R. (1995). Particle swarm optimization. Proceedings of IEEE Internati…

XThread

Corporations solved immortality. Partially. https://t.co/d0yy3KzI4i This is in jest. That said, there’s something amazing about a system where every human can get repeatedly replaced over hundreds of years but the organization keeps compounding. Onchain entities and AI agents will take this much further. Original from here:

XQuote

@karpathy Yes, but we could add an eval for speed. I’m sure it’s a consideration but it could become a more prominent one. “Give the best result you can get in one second of compute on this hardware.” The https://t.co/gI5yz3tgy4 (not grok) guys show what’s possible in terms of speed.

XThread

We could address this via constrained execution. Constrain output length, like the 140 character limit on Twitter. Or constrain the runtime, like real-time mode in Linux. https://t.co/f9nVR7FS7I You don’t need to use the concepts exactly, but ideas from real-time OS development could be baked in as constraints during AI training and evaluation. Soft real-time may be sufficient. Unless it’s a car crash if it doesn’t work, like hard real-time. https://t.co/OpEnybplIy

XQuote

Bitcoin triumphs over the banks. The network triumphs over the state. https://t.co/vti3VzW5sk

XQuote

Perhaps obvious, but... The US government doesn't make money by taxing (like many presume) or by manufacturing (like many desire) but instead by printing. That is, it makes money by administering a global financial empire. And dollar inflation is global taxation.

XThread

THE TELEGRAM NETWORK STATE? The Telegram TON IPO could be important. Mainly because they get the importance. But first...ok, yes, what got announced it isn’t technically a Telegram offering, nor is TON formally issued by Telegram, nor is it really even an IPO. To be precise, what just launched is a crypto treasury company called VERB that holds $500M+ of TON, which is a coin issued by a foundation that was once formally affiliated with Telegram and then became independent but is now affiliated again. Got all that? But these are really just implementation details. The offering could be important[1] simply because (a) Telegram has a billion …

XPost

@sriramk Games, yes. Apps, I don’t know. Games and entertainment can tolerate probabilistic output, so AI can fill screens unpredictably. Apps and businesses often can’t tolerate unpredictability. So they need to bound the use of AI.

XThread

The decline in social trust is a Western phenomenon. And particularly an American phenomenon. Because the opposite is happening abroad. Look at Edelman's trust numbers for China and India. They've been ascending with technology. https://t.co/JnTI2ufDl8 You can of course argue with the Edelman numbers[1] for any given jurisdiction. For example, I am skeptical about their stats for Nigeria and Kenya. However, their numbers do track both (a) the GDP figures[2] and (b) the visible improvement on the ground in places like Saudi Arabia, Malaysia, India, and China. In general, the big picture is that the former first world is now mostly the desce…

XThread

A republic, not an empire. But what does that mean? For Britain, it meant decline. https://t.co/jppo5d9Uds We see this sentiment all the time in tech. A Google that’s nostalgic for the garage. A global behemoth that warmly remembers its womb. And at the scale of a tech startup, it actually is possible to build a giant company and then scale another one from scratch within one human lifetime. But at the scale of a country…it’s harder. Were the US actually so precipitous as to shut down its global empire and return to the proverbial garage, that would likely mean a 99% markdown for US living standards and the end of the dollar as the reserv…

XQuote

Regulate big tech with a thousand startup piranhas. Or protect big tech with a thousand anti-startup regulations. Those are the choices.

XThread

LLMs may have topped out, for now. But the broader AI deployment has just begun. https://t.co/uwGgQ6uxTM Figure is from @tomaspueyo's recent post: https://t.co/M8AOd44H8t https://t.co/tIYBlX37Vw @tomaspueyo It took 20 years from the 2005 DARPA Grand Challenge. And billions of miles, and billions of dollars. But AI can now do this valuable task end-to-end. And probably many others. https://t.co/Ajqk7EhKWr @tomaspueyo Like the transition from artisanship to mass production. https://t.co/uaekJxRV1O

XPost

Introducing AI agents alongside humans on the Internet is like introducing cars alongside humans on physical streets. (1) Some spaces will be AI-only, like bots trading against each other, similar to dedicated highways for cars. (2) Some will be human-only, like proof-of-human social networks, similar to walkable communities offline. (3) And some will be a mix, like Claude Code, similar to city streets with human and machine mixed in close proximity. But the whole digital world will need to be rearchitected.

XThread

There are now two kinds of AI retards. The first kind of retard uses AI everywhere, even where it shouldn’t be used. The second kind of retard sees AI everywhere, even where it isn’t used. Usually, it’s obvious what threads are and aren’t AI-written. But some people can’t tell the difference between normal writing and AI writing. And because they can’t tell the difference, they’ll either overuse AI…or accuse others of using AI! What we actually may need are built-in statistical AI detectors for every public text field. Paste in a URL into an archive.is-like interface and get back the probability that any div on the page is AI-generated. …

XThread

The deep point is: the Internet is not the same as America. Every graph shows that the Internet is economically competitive. But America minus the Internet is not. All the green shoots, all the remaining optimism really comes from the Internet. It’s AI, robotics, drones. It’s people building things and working together and doing well. Unfortunately, much of the non-Internet US economy is not like that. Since the 2008 financial crisis, many non-Internet companies have gradually become zombies, dependent on Fed printing and state intervention both directly and indirectly. The Internet is as American as America was British. America started …

XThread

Good rebuttal to Cloudflare by Perplexity. The core point is that an AI agent is just an extension of a human. So when it makes an HTTP request, it shouldn't be treated like a bot. Functionally, Perplexity is only making this request at the user's direction. @eastdakota What are your thoughts? If users couldn't delegate their actions to AI agents, and all agent traffic was forbidden by robots.txt, then agents wouldn't be able to log in on behalf of users & perform actions. Perhaps robots.txt should get a new section for AI agents.

XThread

I have good news. Yes, the land is moving in an illiberal direction. It’s disorderly anarcho-nationalism and anarcho-communism in America. And disturbingly ordered nationalist communism in China. Businessmen are now randomly murdered in the USA and randomly expropriated in the PRC. All manner of hatred bursts forth every day on social media — against whites, nonwhites, men, women, the young, the old, and everyone in between. And the values of peace, trade, and win/win cooperation that built the modern world seem like they’re on the ropes. But all is not lost. Because the cloud is moving in the opposite direction. As the Internet is actual…

Collection coverage

facebook: unchecked

linkedin: linked

podcast: unchecked

reddit: unchecked

substack: linked

web: linked

x: linked

youtube: linked

Edition 2026-10-10, preview.