0xGaryz

BREAKING: Starlink is now available in the UAE 🇦🇪 Packages start from AED 230 per month Thank you Elon Musk

I packaged up the "autoresearch" project into a new self-contained minimal repo if people would like to play over the weekend. It's basically nanochat LLM training core stripped down to a single-GPU, one file version of ~630 lines of code, then: - the human iterates on the prompt (.md) - the AI agent iterates on the training code (.py) The goal is to engineer your agents to make the fastest research progress indefinitely and without any of your own involvement. In the image, every dot is a complete LLM training run that lasts exactly 5 minutes. The agent works in an autonomous loop on a git feature branch and accumulates git commits to the training script as it finds better settings (of lower validation loss by the end) of the neural network architecture, the optimizer, all the hyperparameters, etc. You can imagine comparing the research progress of different prompts, different agents, etc. github.com/karpathy/autor… Part code, part sci-fi, and a pinch of psychosis :)

Vibe coding is for experienced software devs. Do not vibe code until you really know how to do the basic stuff.

I think it's healthy for us in the Ethereum world to have a more bold and open mindset to many things, particularly on the application layer and on how we see ourselves in the world. We should not compromise on core properties: censorship resistance, open source, privacy, security (CROPS). We should not have "open mindedness" of the type that leaves people with no confidence of what security properties the L1 will still have one year from now. We should not ask ourselves questions like "do we really need light clients to be able to trustlessly verify correctness of the chain?". But especially on the layer of applications and Ethereum's interface to the world, we should be more willing to radically rethink various concepts and step outside our comfort zone. This includes issues of technological direction, eg. "what if AI basically means that wallets as browser extensions and mobile extensions are dead within a year?" One example last year was the shift to thinking about privacy as a first-class consideration, something we value equally to the other types of security. This implies a radically different Ethereum application stack, because the entire stack so far has not been built around privacy. Great, let's build a radically different Ethereum application stack! An example this year is the growing work on the networking side of privacy, both inside the EF and outside. It includes application-layer issues, eg. "what if the rest of defi is basically just universal futures markets on top of a good decentralized oracle and letting users self-organize on top of that?", and "what if the ideal decentralized oracle is just a SNARK over M-of-N small LLMs over zk-TLSes of some major news sites?" (BTW this is interrelated with the AI issue: one consequence of AI is that it moves "applications" away from being discrete categories of behavior with discrete UIs, and more toward being a continuous space, so "build fewer apps and rely on users to self-organize around them" should inevitably expand as a pattern) One example this year is rethinking from zero the role of L2s, and what kind of L2s are actually most synergistic and additive to Ethereum. It also includes culture. This is a big part of "the whole milady thing" for myself, @AyaMiyagotchi and others. Yes, it's a silly meme. Yes, I find the political takes of some milady partisans cringe and sometimes outright bootlickerish (though other milady partisans are quite the opposite). But the core underlying subtext, the message behind the message, is: rip off the suit and tie. If you have your suit and tie on, be willing to grab the nearest wine glass and spill it all over your suit and tie, so you have no choice but to rip it off and reclaim your body's full flexibility and freedom. Actually imagine yourself doing this the next time you get invited to a richpeopleslop formal gala dinner. Take the preconception that you are "respectable", write it down on a piece of paper, crumble it up and burn it. The psychological baptism of doing this leads to the intellectual baptism of unlocking greater creativity and expanding overton windows. For too long, our algorithm in Ethereum has been: we have this existing ecosystem, what's the logical next step to make it one step better? Now, our algorithm should be: we have this L1 that is amazing and will become more amazing, we have a growing array of tools, both those built within our ecosystem and outside it, what are the most valuable things to build, knowing what we know now? If YOU had to write the section of the 2014 Ethereum whitepaper that talked about applications, and take a first-principles perspective of what makes sense in defi, decentralized social, identity, and elsewhere, what would you write? At least take the step of marking all path-dependence concerns down to zero, pretend for a brief moment that the Ethereum chain today has exactly zero usage and you're the one suggesting or building the first apps, and see what comes out. Do this even if you're the one building today's existing apps. This is how Ethereum can grow back stronger.

my artists are cooking as usual 😹😹

.@LinksDAO global membership has been the best IRL value crypto asset of the past four years, and as far as i can tell, it’s not particularly close

3 years ago, you might’ve known @llama_index as a RAG framework. Today we are not a RAG framework. We are an agentic document processing platform 🦙📑 I wrote a blog post detailing the evolution of our company over the past ~3 years and why we believe our current position is enduring in the rapidly evolving landscape of evolving AI. There are two main points that I want to highlight: 1️⃣ One of the most important opportunities in today’s world is to provide high-quality unstructured context to AI agents. We see ourselves as the best in class OCR module that can unlock context from the hardest document containers (PDFs, Word, Powerpoint, Excel, and more) 2️⃣ Agent reasoning loops have gotten a lot more sophisticated. General LLM abstractions are a lot less relevant. Retrieval patterns have completely changed. We need to build deep, focused tooling that actually provides value in this world of long-running agents. Note: We are not giving up on OSS tooling. We think open-source software is extremely important for democratizing AI access. We will continue to build OSS that is more aligned with our core focus area of AI-native document processing. We will continue to support framework users and point them to updated resources for relevant releases. Come check out our blog: llamaindex.ai/blog/llamainde… Our core managed platform is LlamaParse. If you’re interested come check out our platform: cloud.llamaindex.ai/?utm_source=xj…

@chamath Agile makes sense when there’s a clear goal and you assemble a task force for it. Otherwise, you can’t be sprinting all year, for years on end. Hence, the definition of the word sprint.

Agile Has Broken Your Company The Agile Manifesto was signed in 2001 by 17 developers trying to fix broken software projects. It worked…until it didn't. Twenty-five years later, Agile has become a $20B+ industry, and the software it produces is getting worse. The Four Principles The Manifesto prioritized: - Individuals and interactions over processes and tools - Working software over comprehensive documentation - Customer collaboration over contract negotiation - Responding to change over following a plan These aren't wrong in isolation but the problem is what they became in practice. "Responding to change" became an excuse to never finish anything. Stanford researchers found scope creep was institutionalized and rebranded as "sprint replanning," one of the top drivers of cost overruns. "Working software over documentation" quietly gutted institutional knowledge. A 2023 GitLab survey found only 12% of developers felt their codebase was well-documented. In other words, technical debt became structural. "Velocity" replaced quality. Story points. Burn-down charts. Throughput. None of these measure whether the software is any good. The Manifesto said build software that works, and a focus on velocity forgot that. The Numbers Are Damning McKinsey found technical debt now consumes 20–40% of engineering capacity in most large organizations. The Consortium for Information & Software Quality estimated poor software quality cost U.S. companies $2.41 trillion in 2022, with $1.52 trillion from operational failures alone. Agile has been the dominant methodology for most of that period. The Standish Group's CHAOS Report found that in 2020, only 31% of software projects were considered successful. What You Don't Notice Until It's Too Late Current software development best practices have killed systems thinking. When your planning horizon is two weeks, you don't design systems anymore, you assemble features. The result is a mess of fragmented architectures, microservices sprawl, and codebases no single engineer fully understands. The "Product Owner" role that was supposed to represent the customer became a bureaucratic proxy. A layer between engineers and business outcomes, distorting requirements at every handoff. The Alternative: Software Factory The best engineers have always known what actually works. They write specs. They think in systems. They document decisions. They go slow to go fast. At 8090, we call this approach Software Factory. We look at software delivery like a production system with defined inputs, quality gates, and measurable outputs. Architecture is a first-class citizen from day one, not something you refactor into after 40 sprints. Documentation is built in, not bolted on. Quality Is Speed Every hour spent on rework, incident response, and technical debt is an hour that could have gone into upfront design or testing. Speed and quality don’t need to be in tension - it’s a false choice in modern mythology. If your team still measures success in story points and sprint velocity, ask yourself: What's your defect rate? Your documentation coverage? Your time to onboard a new engineer? Your incident frequency? If you don't like the answers, it's probably time for a different model. Try Software Factory at 8090.ai

God's eye view 24-hour replay of Operation Epic Fury. The Iran strikes kicked off and I set an AI agent swarm loose to record every OSINT signal I could find before the caches cleared. Built a full 4D reconstruction in WorldView. I can scrub through minute by minute and watch the whole thing unfold on a 3D globe: > Airspace clearing over Tehran > Ground strike coordinates locking in > Severe GPS interference blinding the region > EO and SAR satellites making passes over the strike zone > No-fly zones locking down 9 countries > Shipping fleets scrambling at the Strait of Hormuz It's pretty amazing how complete of a picture you can build without "proprietary data fusion" -- one dev with public signals and a love for computer graphics and geospatial intelligence. Thank you for all the love on my last post. Dropping WorldView in April. This my friends is just the beginning.

🚀 Introducing the Qwen 3.5 Small Model Series Qwen3.5-0.8B · Qwen3.5-2B · Qwen3.5-4B · Qwen3.5-9B ✨ More intelligence, less compute. These small models are built on the same Qwen3.5 foundation — native multimodal, improved architecture, scaled RL: • 0.8B / 2B → tiny, fast, great for edge device • 4B → a surprisingly strong multimodal base for lightweight agents • 9B → compact, but already closing the gap with much larger models And yes — we’re also releasing the Base models as well. We hope this better supports research, experimentation, and real-world industrial innovation. Hugging Face: huggingface.co/collections/Qw… ModelScope: modelscope.cn/collections/Qw…

we're making @blocks smaller today. here's my note to the company. #### today we're making one of the hardest decisions in the history of our company: we're reducing our organization by nearly half, from over 10,000 people to just under 6,000. that means over 4,000 of you are being asked to leave or entering into consultation. i'll be straight about what's happening, why, and what it means for everyone. first off, if you're one of the people affected, you'll receive your salary for 20 weeks + 1 week per year of tenure, equity vested through the end of may, 6 months of health care, your corporate devices, and $5,000 to put toward whatever you need to help you in this transition (if you’re outside the U.S. you’ll receive similar support but exact details are going to vary based on local requirements). i want you to know that before anything else. everyone will be notified today, whether you're being asked to leave, entering consultation, or asked to stay. we're not making this decision because we're in trouble. our business is strong. gross profit continues to grow, we continue to serve more and more customers, and profitability is improving. but something has changed. we're already seeing that the intelligence tools we’re creating and using, paired with smaller and flatter teams, are enabling a new way of working which fundamentally changes what it means to build and run a company. and that's accelerating rapidly. i had two options: cut gradually over months or years as this shift plays out, or be honest about where we are and act on it now. i chose the latter. repeated rounds of cuts are destructive to morale, to focus, and to the trust that customers and shareholders place in our ability to lead. i'd rather take a hard, clear action now and build from a position we believe in than manage a slow reduction of people toward the same outcome. a smaller company also gives us the space to grow our business the right way, on our own terms, instead of constantly reacting to market pressures. a decision at this scale carries risk. but so does standing still. we've done a full review to determine the roles and people we require to reliably grow the business from here, and we've pressure-tested those decisions from multiple angles. i accept that we may have gotten some of them wrong, and we've built in flexibility to account for that, and do the right thing for our customers. we're not going to just disappear people from slack and email and pretend they were never here. communication channels will stay open through thursday evening (pacific) so everyone can say goodbye properly, and share whatever you wish. i'll also be hosting a live video session to thank everyone at 3:35pm pacific. i know doing it this way might feel awkward. i'd rather it feel awkward and human than efficient and cold. to those of you leaving…i’m grateful for you, and i’m sorry to put you through this. you built what this company is today. that's a fact that i'll honor forever. this decision is not a reflection of what you contributed. you will be a great contributor to any organization going forward. to those staying…i made this decision, and i'll own it. what i'm asking of you is to build with me. we're going to build this company with intelligence at the core of everything we do. how we work, how we create, how we serve our customers. our customers will feel this shift too, and we're going to help them navigate it: towards a future where they can build their own features directly, composed of our capabilities and served through our interfaces. that's what i'm focused on now. expect a note from me tomorrow. jack

China’s response to Anthropic lol

🚨JUST IN: BLACKROCK'S NEW ETHEREUM STAKING ETF TO GIVE INVESTORS 82% OF YIELD @BlackRock has filed an amended S-1 for the iShares Staked Ethereum Trust (ticker: $ETHB), revealing investors will receive 82% of staking rewards. The remaining 18% is split between BlackRock and @coinbase, which serves as the staking execution agent. Between 70-95% of the fund's $ETH will be actively staked, with estimated yields of around 3% annually. Expected to launch in H1 2026.

Introducing: built-in git worktree support for Claude Code Now, agents can run in parallel without interfering with one other. Each agent gets its own worktree and can work independently. The Claude Code Desktop app has had built-in support for worktrees for a while, and now we're bringing it to CLI too. Learn more about worktrees: git-scm.com/docs/git-workt…

Wrote up some notes on the February 2026 update to the official SWE-bench leaderboard, with a bonus side-quest to get Claude for Chrome to redraw their chart to add percentage labels to the bars simonwillison.net/2026/Feb/19/sw…

Coding agents are fundamentally changing software engineering in terms of velocity, role, and org structure. We published a memo to our internal engineering team detailing our growing expectations in terms of role/scope. 🟠 Before, the tasks of prioritization, engineering planning, and implementation were divided between EMs, PMs, senior ICs, and junior ICs 🟢 Now, ICs are expected to handle *all* of product prioritization, product speccing, and implementation This is due to a few trends 📈: - Coding agents have brought implementation costs down to ~0. The role of engineers is writing prompts - LLMs and sub-agents have reduced the PM work of synthesizing feedback down to ~0 too The main job of any “engineer” is to be an e2e product owner: being able to translate requirements into specifications, and delegate tasks to various subagents for implementation. Every engineer is told to offload as much as possible to their favorite tools, whether it’s Claude Code, Cursor, Devin, Codex, regular ChatGPT and more. We celebrate and share learnings around burning tokens, as long as it helps drive additional productivity!

one of the biggest realizations I've had working on Claude Code is that you fundamentally have to design agents for prompt caching first, almost every feature touches on it somehow I wrote this in a day but it's the culmination of months of learnings, hope you enjoy it

Ethereum Is for AI by @DavideCrapis, AI Lead at @ethereumfndn A settlement layer is only as valuable as the amount of economic value that settles on it. Full video below 👇🧵

Dana White vs Eddie Hearn. Boxing's New Power War Begins | M2MMA Podcast... youtu.be/RPwJrBx8pSw?si… via @YouTube Who do you believe will come out on top? @danawhite or @EddieHearn? Either way it’s great news for Boxing! #Boxing #mma #Sports $RLAB

Accurate