Jack Alderson

@natolambert It's the first one unfortunately

So much rests on which of these trend lines is more representative.

@zacamaso @deepseekinyou @yacineMTB You fill out yet own forms?

The db is just to store info and generate context for whatever query i have related to my taxes. I literally ask the AI what numbers i should put it a given tax form and it agentically looks through relevant context and calculates the right answer with advice for being more tax efficient

codex just saved me more in my tax return than i could spend on it in a year

@TheStalwart How are you feeling about crypto now?

HOW THE INTERNET CHANGED IN JUST THE LAST WEEK In today's newsletter, I wrote about how some very straightforward roadblocks to an AI-centric internet are quickly disappearing. This newsletter also includes a Fedlock update that shows Fed hawkishness continuing to surge.

@steipete @cathrynlavery @openclaw Why can't it steer, like codex?

@cathrynlavery @openclaw huh, that's not the default tho?

OpenClaw tip 🦞 One of my most annoying @openclaw paper cuts was when my agent would say “⚠️ Previous run is still shutting down. Please try again in a moment.” Turns out my message queue was set to interrupt, so the new message could collide with run cleanup. Changed it to followup, so the next message queues instead. Tiny config change → massive annoyance reducer 🤗

@0xdoug Those are rookie numbers.. I have 0 technical abilities and am unemployed and I go through ~3 billion tokens/month. (Not anthropic though, fuck those guys)

I’m really struggling to see how the back of the envelope math on this works out… There are generously 4 million characterized “software workers” in America. That’s pretty broad and includes a lot of people who aren’t really classical engineers don’t produce that much code. That comes out to nearly $1k per month of average Claude spend across every dev in America. Yes, there’s some international usage, but it can’t be that much. Yes there is some non software Cowork usage, but that doesn’t use that many tokens. Yes, some non engineers are using Claude to vibe code, but I really doubt many are spending hundreds per month on. Even if we assume 50% of all software workers are using Claude, that comes out to $2k spend per month per Claude user. Thats 10X more than the highest tier Max subscription. So almost all of Anthropics revenue has to be API billing So the only explanation is that something like 20%+ of software engineers are not only Claude users but on API billing and regularly spending thousands per month. At $5/m Opus tokens that means the average API user has to be going through something like 25 million tokens per day. *OR* the other possibility is API revenue is heavily power law dominated. Maybe there’s just something like 100k super users who are making up the majority of the revenue. For that to work the typical super user would have to be spending on the order of $50k/month and guzzling nearly 1 billion tokens per day.

@mert This is not miami

first observation since being in america for first time in a while: supernatural levels of obesity

@itsbrendanma @arbitrum Where are the transactions though?

Three months ago, @arbitrum had roughly 7.8 million stablecoin holders. Today, that number is fast approaching 10 million. +28% growth in under a quarter.

@garrytan Uses obviously ai generated poster for his anti ai ad

Saikat Chakrabarti participated in the creation of tech wealth by being early at Stripe. Now he wants to pull up the ladder on the whole industry. It’s craven. It’s selfish. It is entirely driven by wanting power over us. He is unfit for any public service. Tell your friends.

@VictorTaelin You need a knowledge wiki, like Andre karpathy talks about

seriously, working with AI is MISERABLE for one and only one reason: having to re-explain the same thing "oh yeah this new session obviously doesn't know what proper case trees are, so let me explain it for the 5000th time in my life" I'm tired AGENTS.md doesn't solve this because it is impossible to fit the entire domain knowledge without nuking the context - it would be 1m+ tokens worth RAGs don't solve this, the agent won't search unknown unknowns SKILLs don't solve this unless I keep like a collection of 1750 skills with specific cuts of domain knowledge for each possible subset of my domain that I might need in a given chat, but that's a lot of manual work recursive LLMs or whatever don't solve this for the same reason, you can't dump a domain book and expect the AGENT will magically guess that it is supposed to search for a specific bit knowledge. unknown unknowns fine tuning doesn't solve this (OSS models suck and OpenAI / Anthropic gave up on user fine tuning) I honestly think a good product around fine tuning on your domain would be a major hit and an underdog lab should take this opportunity

I am 95% sure openai has a Mythos tier model, but they were smart enough not to announce it and just used it to create 5.5 etc..

@cremieuxrecueil Why would anyone care? Knowledge retention in the age of AI is like being able to multiply large numbers in their heads in the age of calculators

Neat! Students who did normal studying had greater knowledge retention than students who studied with the help of ChatGPT.

@VictorTaelin @TheRealAdamG Yes exactly. AI needs a promoting skill

@TheRealAdamG yet, if you ask GPT 5.5 to write a prompt for itself, it will do exactly the opposite, i.e., it'll lay out a step by step ultra precise plan on what must be done. you guys need to include some kind of loop that teaches GPT 5.5 how to prompt itself

developers.openai.com/api/docs/guide… **NEW: GPT-5.5 Prompting Guide** "GPT-5.5 works best when prompts define the outcome and leave room for the model to choose an efficient solution path. Compared with earlier models, you can often use shorter, more outcome-oriented prompts: describe what good looks like, what constraints matter, what evidence is available, and what the final answer should contain. Avoid carrying over every instruction from an older prompt stack. Legacy prompts often over-specify the process because earlier models needed more help staying on track. With GPT-5.5, that can add noise, narrow the model’s search space, or lead to overly mechanical answers. For more detail on GPT-5.5 behavior changes, start with the Using GPT-5.5 guide. This guide focuses on prompt changes that follow from those behavior changes. The patterns here are starting points. Adapt them to your product surface, tools, evals, and user experience goals."

@0xngmi Long deserved. Some of the best actors in the space. Congrats

LlamaAI usage is growing 281% Month-on-Month We already had to do 2 migrations in April just to scale capacity

@Youssofal_ @thekitze @sama According to my codexbar, a $200 sub is worth ~$10000 /month of API tokens

@thekitze @sama You don’t understand the scene. The whole AI scene is heavily subsidised and operating at a loss. They lose money on your $200 plan. Your $200 plan is more like $2000 in API credits. But they are subsidising it for user acquisition. They want new users, not existing users.

yo @sama let us buy multiple $200 plans without switching accounts. do you hate money or something??? i'd easily pay $1-2k if i can utilize them properly

@emollick Why not just use AI for the peer reviews?

Very cool analysis of the submissions to a major management journal that shows how much the system of science, built for humans, is under strain as a result of AI. AI can be used to do better science or it can be used to just do more stuff. The danger is that "more" is winning

@cremieuxrecueil 'wise'?

If this was how the buttons looked, what portion of humanity would press blue? It'd probably be a large enough number due to mistakes, the young, altruists, etc., such that it remains wise to press blue.

@TheStalwart Your choice decides whether you get charged conversion by your bank or the merchant pos provider. Your bank will almost always give you a better rate

Question for a payments expert. What I pay for something by card in Spain, the reader offers me the choice to pay in euros or dollars. Obviously dollars are removed from my account. And the cafe gets euros. So what’s really going on in the transaction that requires this step?

@nahuelhilal @Tesla Just hate how it beeps as you if you're not staring at the road.. we all know I'm not going to react faster than the computer, let's just make this legal

Yesterday I drove my @tesla 900 miles on FSD from Miami to Nashville and I realized it’s genuinely the better option. I fly that route 2 to 3 times a month. Flights are never under $400. Most times $600. Sometimes $800. Add Uber to and from both airports, or parking garage fees. Then factor in the delays, the cancellations, the security theater, the chaos, the guy next to you who hasn’t met deodorant yet. On the other hand: I pack healthy snacks, press one button, and the car just goes. I took calls. Replied to emails. FaceTimed my family. Ate without pulling over. Did everything I normally do on a travel day, except none of the stuff that makes travel days miserable. My biggest concern going in was range and charging. Here’s what actually happened: My bladder needed one extra stop the car didn’t even suggest. Most charging stops were under five minutes. Total cost for the whole trip was less than just the uber to the airport. And this was the base model Y. Now I’m thinking I should get something comfier and just make this the default.

@TheStalwart Disagree. Politicians are trying to look hip. Sam is obviously making fun of genz (I am speaking as a fellow millennial)

I can’t quite articulate it, but comms like this reminds me a little bit of Democratic party elected officials

@jessegenet My kid 4 year old can't even watch walleye or Nemo because of the 'scary' parts.. how are they able to watch orcs and goblins?

Watching Lord of the Rings with my 4 and 5 year old and tbh they love it… This is a PSA that you can be liberated from kid movies if you are willing to answer questions every 3 minutes about which bearded guy is which 😅