coldfloat @coldfloatrp

THIS IS STARTING TO FREAK ME OUT I LET GROK COPY 6 WALLETS WHILE I SLEPT. $100 AT MIDNIGHT, $5,463.04 BY 8 AM. I gave Grok Bot one job before bed: copy the best memecoin wallets, and sell when they sell. With one change. When the wallet takes profit, the bot sells 70% with him. The other 30% stays in. I call it the rider. 4 hours in, 27 trades, the balance was $865.07. A stop on $WOOLY -$40.53, and the bot moved on. The money was in the riders. $BRRR: copied $101.76. The wallet sold at 2x, the bot sold 70% with him. The rider ran to 8.3x. +$263.57 $NARWHL: copied $549.45. 70% out at 2x with the wallet, 30% kept riding. 6h 42m into the night it hit 12.7x. +$2,148.36 Most copy bots exit when the wallet exits. Fine, until the wallet sells at 2x and the coin goes 12x without you. Selling 70% at 2x returns 1.4x your entry. The whole position is already in profit. The 30% left is free: if it dies, you still won. If it runs, it pays for the night. Final: 43 wins, 10 losses, 81% win rate, 2,744 tasks across six agents. $100 became $5,463.04. 54.6x. Two riders made $2,411.93. 45% of the profit came from the 30% the wallets left on the table. Paper money, simulated fills. Not financial advice. The full setup is below, free. Bookmark it before tonight.

tenzo @0xTenzo

a month ago

x.com/i/article/2099…

31 39 415 1.2M 1K

@1920web1080 no config no prompts just install. the barrier just died

@gippp69 hourly is fine until you hit that 5 min trap

@issatheblows map first so it doesn't dump the whole repo every time

@semichenkko opus bricks feel like glass. gemini feels like candy

@me_barnyx pixel match against photoshop is the part that actually matters

@irongiantXBT the approval rule is what keeps these from becoming a mess

@kocer_eth this is the real token killer. not the model, the context

@monokern $30k to $0 hit different if the data actually stays live past day 2

I GAVE MY GROK BOT $100 AND ONE RULE: 2 LOSSES AND YOU'RE FIRED. 24 HOURS LATER: $3,662 The first trader it kicked out was the #1 wallet of the week. Firing him cost me $5.95. How it works: Grok Bot takes a trader from the weekly top, copies every buy at 0.25x his size, and sells the second he sells. Paper account, simulated fills. My only job was to watch. - 00:00, hires @mako_sol, #1 of the week - 05:15, two reds in a row, fired. Wallet at $185 - 05:15, hires @nori_ape, #3 - 12:36, two reds in a row, fired. Wallet at $330 - 12:36, hires @velvetbag, #2. At midnight the wallet reads $3,662.88 Then I replayed the same day with no rule. Same $100, same #1 trader, never fired. He kept trading after his two reds and drifted down. That wallet closed at $178.82. So the profit didn't come from picking the best trader. It came from leaving him on time. The #2, the last hire, carried it from $330 to $3,662. Two late sells paid $771 and $1,740. Two firings, $26.89 in losses that triggered them. That's the full price of the rule. 49 copies, 39 green. At 0.25x and out on his sell, the worst red was -$30.38. Hour 24. @velvetbag is still in the seat. One more red pair and he's gone too. Paper money, simulated fills. Not financial advice.

boan @0xboan

2 months ago

x.com/i/article/2092…

104 142 1K 2.7M 4K

I STOPPED CLICKING. IT DIDN'T. I GAVE OPUS 5.5 A PAPER ACCOUNT AND A 16-LEG SCANNER THAT READS WALLETS, NEWS, ON-CHAIN AND SENTIMENT IN ONE PASS. PAPER RUN: $200 INTO $2,946 IN 10 DAYS. A human trader reads one thing, then the next, then sizes the trade. By the time you're done, the move is gone. This reads everything first and only then decides. 8 sections. 340 pairs across crypto, gold and silver. Around 9 scans every second. Every read feeds one score. 21 at the start. 47 after news. 66 after liquidity. 91 before execution. Nothing opens below 72. Risk per trade is capped at 3%. Then it acts. Stop to breakeven at +1R. Take profit at +3R. In one 4-second window it closed SOL, moved a BTC stop and opened gold. $312 after 14 trades. $806 after 31. $1,702 after 47. $2,946 after 58. 69% win rate. I lose a trade every time I check my phone. Claude doesn't have a phone. Paper account, simulated fills. Not financial advice. The full setup is below and free. Almost nobody has it saved. Save it before the next red week.

INSIDER @x_insider4

2 months ago

Your backtest is lying to you and you don't even know it. This is the $200K quant stack — rebuilt with Claude in one evening. Full code. Copy it tonight or keep trading noise. x.com/i/article/2094…

38 85 584 1.7M 3K

@learnodd watched the face first. remote is the whole product. booth buys a kneel, not a shift.

ENGINEERS LEFT THIS ANTHROPIC NOTE IN A CLOSED FOLDER FOR 10 MONTHS, AND THE SCHEMATIC IN IT IS WORTH MORE THAN A $2,000 AGENT COURSE inside is one desk on Opus 5.5 that stops a frontier model sorting the hour every step is a typed decision with a structured output, the cheap model picks skip, summary or escalate, reads the score and sends only the low-confidence pile up on a celebrity tip you feel it immediately, the cheap pass kills it and on a restated revenue footnote a plain prompt just dies: 0.91 against 0.62 on the triage row, and the expensive model stays asleep for the rest what comes out is not a take, it is a memo + the source + a confidence score + a flag a human still opens save this and paste it into Claude Code - let it run the hour through this schematic ↓

Noisy @noisyb0y1

3 weeks ago

x.com/i/article/2101…

12 15 129 344K 233

@shmidtqq sleep is the demo. the gate is the product. money and merges still wait on your yes.

Grok Bot just turned a Polymarket scan into something that actually looks engineered [this setup puts six rules between 7,097 markets and one button] /1 Vash reads every market resolving inside 24h and never skips one /2 Iker reads X live and never trusts a single post /3 Hoss prices the gap and dies anything under 7% /4 Lena sizes the bet at 5% of bankroll, Rudd presses it, no human in the loop /5 Oksa pays the bill from profits, never from capital. daily cost stays $6.58. the screen ends at $3,012.96 the real alpha isn’t a smarter model. it’s a kill line the bot is not allowed to cross. 👇

Gipp 🦅 @gippp69

3 weeks ago

x.com/i/article/2103…

32 11 183 255K 264

ONE GROK FILE RUNS 4 AGENTS FOR $1.80 A DAY AND CLOSES 90% OF THE WORK BEFORE YOU TOUCH THE PRICE a desk that reads, builds, writes and counts costs $55 a month here, the live version of that desk is a salary - a gap you can feel on one invoice 4 agents, 1 file, 0 salaries, and all four read the exact same root Grok 4.5 Fast runs the scout for $0.85 a day, because it reads the feed all day, scores it, and keeps 8 Grok Build 0.1 sits on the maker for $0.35 - the page, the line, the replies - spec, build, live Voice writes in yours for $0.45 a day, angle, draft, queue, the only pass that has to sound like you Ledger on the batch handles the count for $0.15, three numbers, one patch, the cheapest role and the one that rewrites the root 400 scored, 12 written to, 3 answer, 1 pays, and that single number decides what the file was worth the four close everything with a correct answer, someone else picks what to sell and at what price - they can give you the 90%, they cannot give you the business between the four sits one file, and without it these are just four tools in one folder, not a desk you write the root and leave the last 10% on your desk, everything between those two actions is theirs save this and paste it into Grok - let it write out all 4 roles for you ↓

Noisy @noisyb0y1

2 months ago

x.com/i/article/2091…

11 21 176 618K 511

A CHEAP MODEL TRIAGES. A DEEP MODEL THINKS. YOU ONLY PAY WHEN THE CHEAP ONE IS NOT SURE. 60 items cross the desk. 16 get skipped. 21 come back as a summary. 23 wake the expensive pass. ingest → triage → deep read → verify → decide five steps. one loop. the small model never writes the answer. celebrity stock tip, sponsored hot pick, generic wrap, meme chatter. skip. the cheap pass throws those out before a token of the big model moves. fed minutes and a dividend date get a summary. a 10-K, an auditor change, a debt covenant, a footnote that restates revenue. those escalate. the screen says the escalated share sits near a third. the other two thirds never wake the expensive pass. and here is the part the pie chart hides. item 17. triage marked a filing footnote skip. confidence was low. nobody looked. the source check caught it and sent it to the deep model. that miss is the product. not the green count. the quoted file is the same split, just priced. Jev reads at $0.042 per million tokens and writes nothing, so the output costs zero. in the public demo, 100 emails went in, and only the ones under 95% confidence reached Kimi. about 31 hard ones. Kimi's share of that run was about $0.068. the receptionist did the sort. the surgeon only saw the hard cases. Vercel said nearly 13% of paying teams tried it in the first 24 hours. 26% of requests. 1.8% of tokens. that ratio is the whole point. and the honest part of this desk: the clip on screen is a simulation, not a benchmark, and not advice. the footer still says the real cost per item is being measured. a low confidence skip can still be wrong. the rule on the card is the rule. low confidence escalates. you verify in the source. the model does not get the last word on a filing. save this and paste it into Claude Code. let the cheap model triage, and only wake the deep one on the third that earns it.

Noisy @noisyb0y1

3 weeks ago

x.com/i/article/2101…

12 15 129 344K 233

A $3 FILE WHERE THE LOCAL MODEL SORTS THE WHOLE HOUR AND THE FRONTIER MODEL ONLY READS THE AMBER PILE electricity is $3 a month, API is $0 on 80% of ticks, Kimi K3 only wakes on the 20% a laptop cannot finish ingest → triage → deep read → verify → decide → back to the hour Gemma and Llama sit on the machine, Kimi connects when confidence drops, you still open the filing skip the celebrity tip, skip the sponsored pick, summarize the Fed minutes, escalate the 10-K, escalate the auditor change, escalate a restated revenue footnote that is the same split the quoted file sold to law firms and clinics: data stays in the office until the small model is not sure $15,000–$30,000 a month in that post is not a smarter chatbot, it is privacy plus a router one Mac Mini kills a $300 cloud bill in five months, after that the cheap pass is power and here is the honest part of this clip: check 17 marked a footnote SKIP, confidence was low, nobody looked, the verify step sent it up the dashboard is a simulation, the route is not save this and paste it into Claude Code - let it build the receptionist in front of the surgeon ↓

Sprytix @Sprytixl

4 weeks ago

x.com/i/article/2099…

19 30 128 338K 479