arcy

@wtf4uk yapono4ka

🌸

@wtf4uk kotactb

pet ur drun right now! /ᐠ > ˕ <マ ₊˚⊹♡

ADL THE ALBUM.. 3.27.26… 💿

@miketwinks dance

my dream number one ❀ ✿ ✾ aloha ❀ ✿ ✾

@0xwhrrari gud sir

I gave my Claude bot $2,000 on Polymarket and one rule: "maximize profit" It lost for 2 weeks Then $2,000 → $15,240 No predictions, no labels, no "if this, then that" Just reinforcement learning - the bot writes its own rules Most bots use supervised learning Train on history, predict outcomes, hope it holds Mine doesn't predict anything It gets reward for profit and penalty for loss 47 features Price, volume, spread, volatility, sentiment, time to resolution Nobody told it what matters 7 actions Buy YES/NO at 5%, 15%, 30% or hold First 2 weeks - chaos 30-40 trades a day, bleeding money Then something shifted The bot learned to wait, trades dropped to 8-12 per day HOLD went from 30% to 62% It figured out timing on its own, 78% of trades happen in the last 6 hours before resolution It discovered Kelly sizing without knowing Kelly exists High confidence → big position. Low confidence → dust It stopped trading crypto entirely, moved to weather and geopolitics - where edges last longer Nobody programmed any of this Win rate: 64.7% Sharpe: 2.84 Max drawdown: -4.1% 1,680 trades, profit factor: 2.12 Weekly retraining every Sunday, if new policy has worse sharpe - auto rollback Runs on a Mac Mini M2 $47/month The bot doesn't know what's going to happen It just knows how to trade

@0xwhrrari omg sir..

@wtf4uk хохохо

happy woman’s day 𓍢ִ໋🌷͙֒✧˚ ༘ ⋆｡˚♡

@pixnvm корова

cow

@hanakoxbt gudddddddd

Jane Street rolls back your application if you can't spin up Monte Carlo in your head Most traders look at a $100 stock and place a bet. Jane Street interns look at the same stock and simulate everything. 500,000 paths. One pricing equation: ST = S0 × exp((r - σ²/2)T + σ√T × Z) Each path is a possible future. Average discounted payoff across all of them is the price. Converges to Black-Scholes within pennies. No opinions involved. Pure math. Now here's why Jane Street tests this on a whiteboard. They hand you a marker and expect this from memory: d1 = (ln(S/K) + (r + σ²/2)T) / (σ√T) d2 = d1 - σ√T Drift μ vanishes completely from the equation. Option price doesn't care where you think the stock goes. Risk-neutral pricing broke everyone's intuition and Jane Street checks if you understand why. The superday filters hard and fast: > Zetamac below 50 means instant rejection. > Unsolvable problems test your iteration with hints. > Mock trading exposes your runtime under real pressure. > Five rounds. No breaks. No second attempts. Two thirds of their interns came from CS. One third from pure mathematics. Finance degrees almost never survive the screening. The result for those who do: $300K to $500K starting. $1.4M company average. $30M ceiling for star traders. The edge isn't knowing the formula. It's deriving it under pressure while five people watch.

@rei_on_base @base @jessepollak ww

building Rei. AI assistant in your messenger for tasks, notes and daily thinking. not "optimize your time by 69%". more like letting your brain finally stop boiling over. applying to @base batches. demo soon. @base @jessepollak

@pixnvm @diednoreason @opensea ez

helloowwwwwwww chibulka is a 1/1 nft collection on @opensea (0.04 eth) if you want one, dm me :) opensea.io/collection/chi…

@SuccinctLabs some gprove

Next week, we Prove What's Real.

me and my pfp

@wtf4uk изи

@Alucard_eth Touch the grass

Airdrop famrers are not community members, they are extractive parasites that deserve to have their socials and wallets blacklisted by every serious project. I have compiled a blacklist of farmers and popular farmer accounts + wallets. DM me if you are a project and want access.

@pixnvm kak dela