Chris On Code

@8teAPi Did someone say trade? Did someone say latency?

to be clear I meant “going public” as in disclosing trade secrets and emerging into the light.

So Jane Street is going public because obviously they see the future where the model labs compete directly with them in the market. The strategic decision is therefore to become a a specialized infrastructure harness for a future frontier model. Tellingly they point out that the latency constraints mean there is no time for inference at the GPU layer, or agentic tool use at the CPU layer, only reflexive heuristics at the FPGA layer. @yminsky is trying to fend off future model lab competition by making Jane Street indispensable to a future AGI. interesting strategy

@joinstashAI a compounding wiki, for teams, as a company, instead of a skill, pretty cool take hope it takes off

Marco

@dbrinker001 Lol bully ball on em

Just hopped in 400 monsterstack good warmup tourney popped the tag off a fresh white tee blessed best player in the field ezswag hours

@MartinShkreli @devjoshstevens @IgMosqueira @Polymarket I genuinely believe that working for a company that is destroying lives especially the youth was what I was put on this earth to do. Anyways did u guys release a public api yet?

@devjoshstevens @IgMosqueira @Polymarket You genuinely believe in sportsbook gambling 🏈 🎰

This is my 3rd week as VP of Engineering DeFi at @Polymarket , and I'm going to be straight: the traction @Polymarket has seen has massively outpaced our infrastructure, and we haven't done nearly enough to scale to keep up. I hear you, and fixing this is our entire focus. We're a major company now, and we need to engineer like one. Here's exactly what we're doing: - Onchain data latency. We're working on making this near-instant so the experience is incredible. - Chain migration. We need more block space, cheaper gas and much smaller block times so settlement is instant. - Transactions are getting cancelled. We understand this is one of the most frustrating issues right now, and we have a complete fix coming very soon. - Massive focus on the website to make it faster, more responsive, and with better UX. - We added observability everywhere. Proper alerting so we catch issues ourselves, market makers should not be the ones telling us something is down. That's been unacceptable, and we know it. - E2e tests throughout, starting with the CLOB, so issues get caught in CI before anything ships. - CLOBv2 is not a rewrite. It won't improve performance or stability on its own; it's an upgrade that unlocks us to move fast right after. We'll do better with communication next time. - We are rebuilding the CLOB from the ground up. Most important thing we're doing. Without it, we can't be the best DeFi exchange in the world. We know it, we're on it, it's mission critical. - Unified TypeScript SDK for all APIs, which is shipping soon. - Unified API. One WS connection for everything, with a schema that's actually readable. - New Polymarket contract in the works that unlocks things that are simply impossible on the current protocol. - New hires: Head of QA Automation, Head of Dev Tooling, Head of Internal Tooling, Head of Data Engineering. - Smaller, dedicated teams. Fewer focus points per person, clearer ownership. People do what they're good at and are accountable for it. - Working closely with customer support to give them real debugging tools so any user issue gets properly diagnosed, not lost. - Proper communication with marketing and market makers so everyone knows what's coming and when, and MM can submit feature requests with a clear path to get them into engineering and shipped. - Working with 4 security teams daily to ensure we're super secure and that funds are always safe. - Perps incoming. Brand new contracts and a backend built from scratch in Rust. We're proud of this one. - A lot of other fixes are running in parallel right now. Starting next Friday, I will be posting weekly engineering updates. I joined because I genuinely believe in what @Polymarket is trying to do. @shayne_coplan built this so the world has somewhere to go to find out what's actually going to happen, not what the media thinks, not what a pundit says, but what thousands of people are willing to put money on. But right now, our engineering isn't living up to that. We've let people down, and I'm not going to dress that up. I came here to fix it, and that's exactly what we're going to do. The next few months are going to speak for themselves. Stay with us.

@arsham_manukyan @elliotarledge This project took me 5 months to build but imma go to bed while it finishes these 25 backend tasks

@elliotarledge I really wonder what the overnight session is? Do you guys just leave the coding agent with the code alone and go to sleep?

just woke up to opus 4.7 nuking one of my projects during an overnight session. luckily i was able to get it back easily

Elon Musk was asked why his companies move faster than anyone else. His answer: "I'm constantly addressing the limiting factor. Whatever the limiting factor is on speed, I'm going to tackle that. If capital is the limiting factor, I'll solve for capital. If it's not the limiting factor, I'll solve for something else." He then said something most managers never figure out: "If something is going really well and making good progress, there's no point in me spending time on it." "The irony is if something's going really well, they don't see much of me. But if something is the limiting factor, they'll see a lot of me." He spends his time entirely on whatever is blocking the next step. Not on what's interesting. Not on what he's best at. But on whatever is the bottleneck right now. Most leaders do the opposite... They gravitate toward what they're comfortable with and away from the hard problem. From: @dwarkesh_sp and @collision

AI has gone too far. Harry Potter "6 7" platform. Credit: Unhindered Studios

@BullyEsq @ledgerstatus Learn to operate Claude in the terminal cli This will lead to learning how to operate terminals Before you know it you’re a dev

What are you guys using for model routing and optimization? Currently plugged into Anthropic+OpenAI+OpenRouter combo but I’m sure people have solved this better already

@aakashgupta All this is level 2 imo but it’s good advice ;) wonder if it’s a human or bot account 👀

There are 6 levels of making Claude Code run autonomously, and most people are stuck on Level 1. Level 1: Kill the permission prompts. Run claude --dangerously-skip-permissions. One flag. Now it stops asking “Can I edit this file?” every 30 seconds while you’re checking Slack. Level 2: Context window management. Claude Code now supports 1M tokens. Use /clear between tasks. Run /compact at 60% usage instead of waiting for auto-compaction to fire at 90% when the model is already forgetting your instructions. Level 3: Subagents. The reason it stops at 15 minutes: everything runs in one context window. Subagents run in separate contexts. Build a looping todo command, each task executes in its own window. Builds, tests, and git operations never touch the main conversation. 2+ hours autonomous with zero intervention. Level 4: Ralph Wiggum loop. Official Anthropic plugin. Claude works, tries to exit, a Stop hook blocks the exit, re-feeds the same prompt. Each iteration sees modified files and git history from previous runs. One developer ran 27 hours straight, 84 tasks completed. Geoffrey Huntley ran one for three months and built a programming language with a working LLVM compiler. Level 5: Karpathy’s AutoResearch. On March 7, Karpathy pushed a 630-line script to GitHub and went to sleep. Woke up to 100+ ML experiments completed overnight. 25K stars in five days. The difference from Ralph: structured eval loops. Define a metric, run, measure, analyze failures, improve, repeat. One Claude Code port took model accuracy from 0.44 to 0.78 R² across 22 autonomous experiments. Level 6: VPS + OpenClaw for 24/7. Your laptop lid closing kills everything. Run Claude Code on a VPS inside tmux. Detach, close your laptop, come back tomorrow to a finished diff. OpenClaw (247K GitHub stars) takes it further: a persistent gateway connecting LLMs to your real tools, running 24/7 across messaging, email, git, and calendars. Jensen Huang at GTC called it “probably the most important release of software ever.” The unlock at every level is the same: give Claude a way to verify its own work.

@aakashgupta What do you mean by looping to do? That’s what ralph is doing.

@lowstakesdre But u could of got it all

Not only does he hate folding. He loves raising. His PFR is higher than some of the nits on this site VPIP. God bless Benny26. A true legend. 🫡

@ImKadaddle LOL for real. Online time- 20,000 hours. Campaign- 5 hours

@dbrinker001 The blacker the berry the sweeter the juice

One of the most underrated skills in poker is the ability play your A game in very soft pools Most pros "Let their guard down" or get too relaxed in the softest games because they know its a +EV spot regardless if they make mistakes or not Good pros get MAXIMUM VALUE

@MonkerGuy @bp22 Claude is better at planning if you feed “it,him, her” garrytan’s planning skills

@bp22 Have you tried codex 5.3? It's better than opus for coding gpt 5.4 high is significantly better at some things, planning, solving problems, but sometimes over complicates things I recommend: Start with 5.4 high Use codex 5.3 for mission critical updates Opus for 2nd opinion

Claude code addiction is a real thing and some of us need help. Speaking for a friend 😜

We’re not in a simulation we’re making simulations 🙀

Once in a lifetime experience

@BlackRainPoker Only if you’re not friendly imo

The best poker players in the world have more haters than anybody else. People won't like you when you keep taking their money.

@TukiFromKL Prove it 👀

🚨 Cursor just dropped Composer 2.. their own AI model.. not Claude.. not GPT.. their own.. and it beats Claude Opus on coding benchmarks.. at a fraction of the cost.. a code editor with 50 people just outperformed a $30 billion AI lab.. at coding.. which is supposed to be their whole thing.. the vibe coding era just got an upgrade..

@dbrinker001 who has a higher roi than you in online wp events? king of the castle

pulled a gold bounty for the boys