Paul

@Atenov_D The architecture is smart. But most pipelines miss the calibration layer: comparing AI probability estimates to actual resolution rates over time. Without feedback loops, the hallucination problem just moves upstream.

@cryptorover Coinbase Premium at -0.079 means US institutions are selling harder than global markets. Last time this divergence sustained this long was March 2024, right before BTC dropped from $70K to $56K.

🚨WARNING: Coinbase institutions are non-stop dumping Bitcoin.

@Cointelegraph XRP inflows at $2.66M vs BTC outflows at $296M is a 1:111 ratio. This is new retail entering a recently ETF-listed asset, not rotation. Institutions are clearly reducing exposure. The dog is still BTC.

🇺🇸 ETF FLOWS: XRP spot ETFs saw net inflows last week, while BTC, ETH and SOL spot ETFs saw net outflows. BTC: -$296.18M ETH: -$206.58M SOL: -$4.24M XRP: $2.66M

@pankajkumar_dev 1M token context != 1M usable context. Coherence degrades around 200k tokens, known attention behavior, not a Claude-specific bug. The problem is Anthropic hasn't published effective context guidelines. Advertised vs. reliable differ by 5x.

Claude Quota Reduced: Opus Feels Worse and Pro Plan is Unusable - I noticed a clear drop in Claude’s quality, especially with Opus 4.6. more hallucinations, worse decisions, and poor instruction following. - Tasks that used to work smoothly now need multiple retries, even for simple changes. - The quota issue is worse even small prompts are consuming a noticeable percentage of limits. - Claude Pro is where things break a single maxed Sonnet session took 8% of my weekly limit. - In just 2 days, I hit 56% usage with only 5-6 sessions (2 hours each). - This is with normal usage like concepts, strategy, and documentation. - Paying $20/month now feels like getting fewer messages than the free plan. - The Pro limit feels almost unusable. - Later it was said peak hours burn quota faster, but this wasn't communicated clearly. - Overall, it feels like Claude got worse and more expensive at the same time.

@RoundtableSpace Scanning wallets post-trade is survivorship bias. The real edge is identifying which wallets historically resolve before the crowd and front-running their next entry. That's a strategy worth building.

CLAUDE TURNED ONE ARTICLE ABOUT A POLYMARKET EXPLOIT INTO AN AUTOMATED MONEY MACHINE. It scanned 349K wallets, found the one that already knew, and flipped $800 into $2,600 in 2 days.

@zostaff This is HFT logic applied to prediction markets. The dead market and arbitrage bots will work until liquidity providers reprice. When the spread widens to absorb the edge, what's the fallback?

5 BOTS. 5,602 TRADES. $22,000 IN ONE SESSION. HERE'S WHAT EACH ONE DOES. 71.6% win rate across the fleet. Arbitrage - finds broken logic between linked markets. One market says 40%, a related one says 25%. Mathematically impossible. Bot buys. +$5,273, 72.7% WR. Calibration - XGBoost trained on all of Polymarket history. Low-probability events are underpriced, markets near deadline don't move to 0 or 1 fast enough, favorites are overpriced. +$5,332, 72.9% WR. News reactor - LLM reads headlines and scores impact on specific markets in milliseconds. Fed speaker drops a hint - bot is already in while people are still reading the headline. +$4,134, 72.8% WR. Dead markets - finds where the outcome is already known but the price is stuck at 92 instead of 99. Picks up forgotten money. +$3,810, 70.8% WR. Order-flow - autoencoder on the live orderbook. Pattern breaks - someone informed just entered. Bot follows. +$3,499, 68.9% WR. Average profit $3-10 per trade. Small edge, thousands of times. They don't predict. They pick up what the market already lost.

@KobeissiLetter Worst 12-week start since Dot Com bubble. S&P futures turn green Sunday night. That's short positioning exhaustion, not a new trend. Until cash session confirms, it's noise.

BREAKING: S&P 500 futures erase all losses and turn green.

OctoBot 2.1.1 has just been released. It fixes issues with @HyperliquidX, configuration of the GPT interface and @Polymarket tickers issues.

@QuintenFrancois The correlation held when central banks were the dominant variable. It broke when fiscal/tariff policy took over. Correlations don't die, they go dormant until the regime returns.

Remember when we all 100% believed the Bitcoin - global liquidity correlation?

@Shelpid_WI3M 1 cent on Polymarket implies ~1% probability. If the true rate is 3-5%, you have a 3-5x Kelly edge. Works until enough bots flood it. How long until 1 cent becomes 3 cents?

Claude gave me an answer I wasn’t supposed to hear. I asked Claude: I have $300. Where do I put it to make the most money possible? Expected the usual. Stocks. ETFs. Maybe real estate. Instead Claude said: Go to GitHub. Search polymarket-1-cent-bot. Find the repo with the most stars. Deploy it with your $300. Most positions will expire at zero. The ones that hit pay 60x. Someone is already running this at scale. Search silent-marmot. I searched GitHub first. Found a 290 star repo. Last commit 2 days ago. 39 lines of Python. README had one line: buys noise at 1 cent, sells certainty at $1. Then I searched the wallet. silent-marmot. $512,430 profit. 31,882 trades. Joined November 2025. Bio: still early. Wallet: t.me/KreoPolyBot?st… I asked Claude: how does the math work? $300 across 30 positions at 1 cent each. 27 expire worthless. You lose $270. Three hit. Each pays $60. You walk away with $180 net profit. If four hit you double. Five and you’re silent-marmot. I put in $120. Copied the repo. Went to sleep. One hit. $120 → $780. I asked Claude: why didn’t you just say buy an index fund? You asked for the most money possible. Not the most comfortable answer. GitHub repo still public. Wallet still running. Bio still says still early.

@Polymarket Polymarket prices $200 oil at 14% before July. But $110 by June is sitting at 70%. The $200 headline is the noise. 70% on $110 is the trade.

BREAKING: U.S. officials & Wall Street analysts are now reportedly preparing for the possibility of $200 oil.

@SherryYanJiang @MiniMax_AI 95% cheaper matters most in agentic loops, not one-shots. A 10M-token coding run costs $150 on Opus vs $8 on MiniMax. Cost savings compound per iteration, not per task. How does it hold on multi-file refactors?

pretty impressed that @MiniMax_AI 2.7 can one-shot a linear clone in 10 mins at 95% cheaper than claude opus 4.6 not enough people know about this model. if you're just using whatever your timeline is hyping, you're probably overpaying for the same result

@hypurrdash The survivorship bias issue is real. What's the 30-day turnover rate of the Hot 100? If top spots rotate fast, the score is descriptive not predictive. The edge is identifying consistency before it shows up in the metrics.

Introducing the Copytrading Hot 100 🔥 Copytrading Hot 100 is a real-time ranking of the most consistently profitable traders on Hyperliquid. We've computed a score for each trader that takes into low drawdowns, high R/R trades, high sharpe ratios and more.

@howdymerry The execution metrics are the real moat. Polymarket spreads on binary markets can run 3-5% wide - a strategy that backtests at +8% EV can go negative in live trading purely on fill quality. How are you modeling liquidity in the sim?

I took Karpathy's Autoresearch concept and adapted it into AutoPredict: a research framework for evaluating, backtesting, and iteratively improving prediction market trading agents AutoPredict evaluates agents on - forecast quality - calibration - execution (slippage, liquidity, and fills) - drawdown and risk adjusted returns It also supports domain specialists for weather, finance, and politics under a shared evaluation harness This framework is NOT for building agents but for agent improvement via a evaluation + mutation + selection loop

@NoToDigitalID The same institution pushing mandatory EU Digital Identity Wallets for all 450M citizens just lost 350GB from their own systems. Centralization doesn't just raise stakes — it multiplies them.

🚨BREAKING: European Commission confirms its website was breached after a hacker said they stole more than 350GB of data. The hacker plans to publish it online.

@glassnode 500ms block time doesn't erase the edge — within-block ordering still favors receipt time. Tokyo's 16ms vs Amsterdam's 245ms is a 14x queue advantage when multiple arb orders hit the same cycle.

Tokyo is pinging the Hyperliquid API in ~3ms. Amsterdam is sitting at ~221ms. Distance is a tax on your execution. We just deployed a live map of global probes tracking API and direct validator latency to Hyperliquid in real-time: glassno.de/4t7kUhv 🇯🇵 Tokyo: ~15.9ms 🇰🇷 Seoul: ~50.2ms 🇭🇰 Hong Kong: ~66.9ms 🇸🇬 Singapore: ~136.1ms 🇺🇸 Virginia: ~163.5ms 🇳🇱 Amsterdam: ~245.2ms

@bridgebench The speed critique assumes parallel agentic workflows. For deep, long-horizon tasks (debugging over days), 44 tok/s matters less than quality. The real question is whether GLM-5.1 maintains coherence at 128K+ context.

GLM 5.1 is the slowest frontier model we've ever benchmarked on BridgeBench. 44.3 tokens per second. Half the speed of GPT 5.4. Nearly 6x slower than Grok 4.20. Z.ai traded all of their speed for intelligence. The coding benchmarks improved. The throughput collapsed. In 2026, agentic coding is about parallelism. You're running 5, 10, 15 agents at once. A model this slow bottlenecks every workflow it touches. Intelligence without speed is a luxury most vibe coders can't afford. bridgebench.ai

@louszbd The coding eval uses Claude Code as the harness, which biases toward Claude's scaffolding strengths. GLM-5.1 at 95% of Claude Opus 4.6 on a Claude-native harness is actually impressive. What does it look like on a neutral scaffold?

finally glm-5.1 at the very beginning we were teaching models how to write code, basically training a system that could imitate developers. back then AI lived inside the IDE as an intelligent assistant, but we were still the main driver. that was the copilot era of AI coding. then it started to become something more collaborative. we could express a vague intention (prompt), and the model translates that intention into structured software. in a way, that was the first time we taught machines to understand vibe. earlier this year, we entered the agentic engineering era. we stopped programming line by line. models began to form plans, maintain them, and operate inside a feedback loop. the model takes responsibility for planning. and now we are approaching a moment where AI can operate on the same time horizon as engineers. this is why we built glm-5.1. we want to unlock a new long-horizon paradigm. where it starts to tackle the kinds of problems that unfold over weeks: debugging, integration. an agent to remember context over long stretches, still stay aligned with the objective (and keep correcting itself along the way)

@0xSero The GPT-5.3-Codex tool drop via BYOK is a pattern, not a bug. Third-party APIs lose reliability at extended horizons. That Claude Opus does consistent 20h runs is becoming a moat on its own. Reliability at scale separates frontier from commodity.

In the last 3 weeks: 1. They've updated the desktop app: - Project folders - Session search - Better support for BYOK - Skills viewer + their skills are top notch 2. BYOK has gotten much better - missions now accessible - you don't ever have to use any of their models - spec mode is phenomenal - Claude Opus can now do 20 hour runs ------ What's still broken: 1. GPT-5.3-Codex via BYOK drops tool calls and just stops working 2. Lots of flickering in Zed Overall happy customer. x.com/0xSero/status/…

@aicodeking @Zai_org Worth noting the benchmark is Kilo Code-specific. Models tuned for one agentic scaffold overfit to it. Claude Opus 4.0 at #1 makes sense on general capability, but the gap narrows on neutral evals. What does GLM-5.1 look like on SWE-bench or GAIA?

GLM-5.1 by @Zai_org is one of the best agentic models out there. I've been testing it early and it is genuienly impressive. Way better at instruction following, long running tasks than previous generation. full review here: youtu.be/UxGieu7PaPg