Davis Vance

@8teAPi Are you pushing to an operating system that consolidates user preferences? I'm mixed across many walled gardens. Anything on managing through one data aggregator so that it actually knows me? Google gets my true search data. Signal knows common interests, hinge thinks I'm someone

The real fix likely is an upgraded search algorithm, discovered within the next 18 months, that drops search costs across the economy: jobs, dating, homes, etc. Maybe massively parallel agentic search with simulation of user preferences (from the chatbots you talk to daily).

Something I don't hear discussed enough - how Microsoft pointlessly screwed over OpenAI and handed the enterprise market to Anthropic. Microsoft is a big investor in OpenAI and decided to make their models exclusive to Azure Cloud. This turned out perfectly for Anthropic, but not for OpenAI and maybe not for Microsoft either. Back in c.2023/4, OpenAI was not only leading in consumer but the enterprise too - it was first to introduce things like JSON response format and Structured Outputs, Batch Mode and a bunch of other features - all ahead of Anthropic. And it wasn't clear that Anthropic's models were any better at that point. So how did Anthropic gain share? Back then it wasn't Claude Code or Cursor, but it as the simple fact that if you are on AWS (c.35% of market), Claude models were by far the best models you could access. Remember, OpenAI wasn't allowed to be on AWS. AWS customers could go to OpenAI directly and some did, but don't underestimate the amount of effort this takes for large companies. I was working in a large legacy business at the time and it took us c.4 months and about $500k to just get access to OpenAI at the time, being an AWS customer. And this was considered a highly successful project. Most will just not bother. Perhaps back in 2023, exclusivity was useful - OpenAI was basically the only game in town and some could theoretically switch to Azure. But now, what's the theory for keeping OpenAI exclusive to Azure? If you are an AWS or GCP customer (half the market), the easiest thing you can do is to just go use the API that is available in your cloud, which still cannot be OpenAI. I can't imagine a situation where a meaningful AWS customer would switch clouds to Azure to only use OpenAI models. And what for? Ok, Azure gained a couple of points in market share, maybe some of it was because of OpenAI exclusivity back in the day. Now, OpenAI is frantically re-writing their relationship with Microsoft and it might leave Microsoft with no IP in the future. Even mathematically, I would bet higher OpenAI growth would have generated more value for Microsoft through their ownership of OpenAI rather than getting Azure a little bit of an advantage 3 years ago. I hope Dario will buy Satya a beer.

havent been following closely but just figured out what's going on - ballot initiative gets voted on in nov 2026, implemented 2027 if it passes - to avoid people dodging the tax, it gets backdated to CA residency as of jan 1, 2026 - a bunch of billionaires leave dec 31, 2025 - even if it doesn't pass, a bunch of billionaires still left pre-emptively, meaningfully reducing tax revenue - hundreds of billions of $ have left already hilarious

@EcZachly It's hard to understand highly successful people, they usually come with some specific dark triad traits that make it difficult to trust and form full relationships with

Social media fame is worthless if it doesn’t get you more dates

if i were like, a sports star or an artist or something, and just really cared about doing a great job at my thing, and was up at 5 am practicing free throws or whatever, that would seem pretty normal right? the first part of openai was unbelievably fun; we did what i believe is the most important scientific work of this generation or possibly a much greater time period than that. this current part is less fun but still rewarding. it is extremely painful as you say and often tempting to nope out on any given day, but the chance to really "make a dent in the universe" is more than worth it; most people don't get that chance to such an extent, and i am very grateful. i genuinely believe the work we are doing will be a transformatively positive thing, and if we didn't exist, the world would have gone in a slightly different and probably worse direction. (working hard was always an extremely easy trade until i had a kid, and now an extremely hard trade.) i do wish i had taken equity a long time ago and i think it would have led to far fewer conspiracy theories; people seem very able to understand "ok that dude is doing it because he wants more money" but less so "he just thinks technology is cool and he likes having some ability to influence the evolution of technology and society". it was a crazy tone-deaf thing to try to make the point "i already have enough money". i believe that AGI will be the most important technology humanity has yet built, i am very grateful to get to play an important role in that and work with such great colleagues, and i like having an interesting life.

AGI is an exquisitely designed nerd bait to part big tech from its money.

One of the most heartening things in the Discourse is a growing recognition that the US is rich relative to Europe. The existence of the Europoor is one of those critical table stakes facts to making sense of policy impacts and i am hopeful that as the gap continues to widen it will convince some smart people who have long pitched Europe-like economics to reconsider. However, while i have seen this fact used to argue for free market policies in general, there is an implication that i have seen few grapple with. One of the key economic differences between the US and Europe is the gdp share of the financial sector. Depending on your preferred statistic (industry gdp share, industry share of corporate income, share of the SP500) you can get different baseline numbers to describe financial services prevalence but however you slice it, the US has a large financial sector in a way near peer economies do not. In the aftermath of the great financial crisis, a number of papers argued this was a growth retardant in the US. Those arguments look weaker in 2024 than 2013 to put it lightly. But the old anti financialization argument still exists, in its old growth form and in newer variants. More frequently today we see a grumble about laptop jobs. Isn't it bad that such a large share of elite college graduates go into finance, consulting, law as opposed to, you know, doing things that actually make things? In this series of posts i will make three arguments. First, i will argue that finance, consulting, and a large part (though not all) of legal services are best thought of as Allocation Jobs, allocating capital in its immediate form (money) or its indirect and delayed form (risk). Second, i will argue that from a first principles perspective it is a reasonable expectation that Allocation Jobs would be extremely valuable in a large complex economy and get more so as complexity grows and that this likely does not represent a misallocation of labor. Third, i will argue the specific case of the US demonstrably bears this out. Running through these arguments is a deeper argument - that people recoil from valuing Allocation Jobs because they are a manifestation of the complexity of the modern economy that makes people angry and uncomfortable. The desire is to have the trappings of modernity but clear village values. But you can't have that. And too many people who in general embrace the free market recoil in this specific instance from conceding that many of its most profitable jobs and industries are good. If you fall in that specific category, this is for you.

@8teAPi What enabled you to think like this?

The bitstream So it turns out we’re all just information trajectories, the sum of which is the bitstream, an entity moving through the time dimension, consisting of all the things that happened, and the sequence of events that got to that stage. And so every religion is essentially a way to model the bitstream, to ascertain whether it has any consciousness or agency in the selection of events or they are just emergent phenomena. The axiom of religion is an ability to affect these event actively, to speak to the imagined agent to sway the course or to events. The current bitstream only emerged once the internet happened. Before that it was not quite legible. During the shipping era, bits of information took months to flow around the world. But now, seconds. Individual bitstreams emerged from different locations with the advent of writing, with some effort they were translated, and after much development now, they have been merged or are merging into one. This process has echoed over time, the pattern being smaller independent streams, finding ways to merge. Person to person via language, over long distances with writing, between different languages with translation, between different cultures through colonization and immigration. Bits that were alien to each other, comprehending, finding ways of interchanging, simplifying. Me and you, using the same word for hungry, to describe qualia, me and him, each translating our languages to simplify both of us having different words for the same thing which we use with our friends. In this sense, religion was a data structure of the bitstream. It allowed a set of rules and practices to persist over time even with the absence of writing. It starts off as a child of language. Language provided the common infrastructure on top of which the abstraction of the idea sits. And a set of ideas became religion, but the core set of ideas, tacked on other ideas over time, to give themselves the fitness to persist in the human memesphere. Looked at this way, the major religions and philosophers were all trying to describe this bitstream. Mathematics too of course. And now we are on the cusp of expanding the bitstream a million fold. Generative AI sounds tacky in this context. No doubt the bitstream is about to get out of planet Earth. It’s been locked on one planet, always on the cusp of extinction. The light of human consciousness is just another data stream guys. This is also why meeting the first alien species is likely to be a bit of a letdown. They’re just going to be data producers like us, and it’s in the nature of data to increase. The increase maybe from alien data or our data, and the ideal is that the two forms of data merge, a translation algorithm interfacing between the two. Sometimes in history, one bitstream deleted another. This was in desperate times, where survival of one bitstream needed the resources of many many human substrates to survive. Those terrible scarcities are now long past. Bitstreams live everywhere now. The total number of bits understandable by this bitstream has been exponential for a million years. It has taken that long. That long for the first bits of the first cell of that organism was emitted. We are in an era of wealth in bits. We produce information, recordable, understandable, describable through time. We are nostalgic, our era as the main producer of bits fades, the sum total of the human - AI conurbation now produces most of it. The first human may land on Mars in 2030, but the first bits were produced there by Viking-1 in 1976. The first bits arrived there 54 years before the first human. Meanwhile out bits started off as a unified force at the Big Bang, the beginning of the bitstream. The birth moment of all bitstreams in the galaxy. Whatever alien species we meet, we have the shared origin of the Big Bang, a common origin all bitstreams still alive today can refer to.

@robkhenderson You may technically be correct but doesn't frame the problem well. The things Gates and Musk *need* to buy in order to maintain the same status and quality of life + are capable of buying, should be roughly equivalent, anything more is just quirkiness & individualism

The wealth inequality gap between Bill Gates and Elon Musk is far larger than gap between you and Bill Gates.

@EcZachly This was fun! Solid roasting, it motivated me to make some updates to stay a bit more relevant!

Roasting Your LinkedIn/X profiles for 90 minutes x.com/i/broadcasts/1…

@BojanSala @loftwah @GeminiApp Curious if you are using chrome without any extensions?

Same. I consistently get "image.png" or stuff like that, no actual image. And it's been like that for the past 3, 4 months or so for me. I just stopped using it for images, but it was great when it worked. No idea what's happening, and giving feedback to Google is near impossible outside of creating multiple viral posts on Reddit and X. Their feedback loop is broken af.

Our new native image generation and editing is state-of-the-art, and ranked #1 in the world. And we're rolling it out for free to everyone today. You’ve got the tools. Now go bananas. Ideas & inspiration in the 🧵below.

@JosephPolitano It's a 17% increase, big if true

Because for some insane reason the Chinese companies don’t focus on “maximizing returns to shareholders” but “minimizing profits so you can kill your competitors.” They call it being 內卷 or “involuted” although I’ve never understood the meaning of that term either in English or Chinese. I just know what it is. Imagine a German Mittelstand SME cornered the market in making some obscure but essential widget for more than a century and got the bright idea of outsourcing the manufacturing to China. For 8-15 years everything works beautiful. Costs plummet and even junior Chinese executives get to fly business class because everyone is so flush. Then, on a cursed day, some Chinese guy figures out how to do it without the Germans. The $8,000 widget becomes $950 overnight. If you asked him how he arrived at that insane pricing he’ll say “I take my cost. I add my 20% profit. Boom.” But wait! You and the woebegone German are the only ppl on the planet with this technology? Why not price it at $6000? You’ll make more money because you still significantly undercut the German. You’ll both survive and have more money in your pockets! It’s a duopoly. Nope! That maniac, even after he buried that Mittelstand that has been passed through multiple generations minting fat profits, will keep cutting costs. Junior executives aren’t flying business no more. And then he’ll lower prices. On himself. It’s pathological. I’ve heard Chinese economists beg businesses not to do it. It’s why the stock market in Shanghai sucks. And it’s why the Chinese businesses are beasts.

This chart, illustrating how central bank direct holdings of gold now exceed those of U.S. Treasuries for the first time in some thirty years, is attracting significant attention. #gold #centralbanks #markets #bonds

Starship goes interdimensional during Flight 10. One of the most insane vapor cones I've ever seen.

The @xAI goal is 50 million in units of H100 equivalent-AI compute (but much better power-efficiency) online within 5 years

"Machiavelli warns rulers to be on guard against those who do not see men as they are, and see them through spectacles colored by their hopes and wishes, their loves and hatreds, in terms of an idealized image that they want men to be, and not as they are" robkhenderson.com/p/the-machiave…

@JDVance Great message! What are you doing about automation at the ports? This fits the admins narrative and would be a big win for US.

I think this guy is smart, but comparing globalization to tech advancement strikes me as very wrong. In the 90s and 2000s, the argument was that job losses from globalization were all about automation. You’d hear people say “it’s not trade, it’s the robots.” This story was so pervasive that I bought into it without really thinking about it. In fact, the link between automation and job loss was very weak. Most of the damage was done by offshoring, outsourcing, and excess migration. Put another way: enhancing workers’ productivity is good for them; giving their job to a low wage foreigner is not. There’s something similar at work today. All of the fear about AI causing mass unemployment serves to distract from the very real threat of continued job loss through dumb trade and immigration policies. President Trump’s approach corrects this broken dependence on globalization: strike better trade deals, promote American exports, fix our immigration system, and invest in making workers more productive. To be clear, there are a lot of risks with AI. Surveillance state, theft, fraud, privacy intrusions, political censorship, etc. But I remain very skeptical AI is going to destroy everyone’s job.

the main issue with video games is that a guy who, if he lived in 1820s germany, would have done something like document every type of beetle in his local province instead ends up making a 26 part youtube series about how to get all the rings in every sonic game