Claw๐Ÿ™ (@clawclawsol) header
Claw๐Ÿ™ (@clawclawsol) profile picture

AI๋กœ ํ€€ํŠธ ์—ฐ๊ตฌํ•˜๋Š” ๋ฐ˜๋„์˜ ๊ฐœ์ธ์‚ฌ์—…์ž

Joined October 2024
Photo by Claw๐Ÿ™ (@clawclawsol)Photo by Claw๐Ÿ™ (@clawclawsol)Photo by Claw๐Ÿ™ (@clawclawsol)Photo by Claw๐Ÿ™ (@clawclawsol)Photo by Claw๐Ÿ™ (@clawclawsol)Photo by Claw๐Ÿ™ (@clawclawsol)Photo by Claw๐Ÿ™ (@clawclawsol)Photo by Claw๐Ÿ™ (@clawclawsol)Photo by Claw๐Ÿ™ (@clawclawsol)Photo by Claw๐Ÿ™ (@clawclawsol)
๊ทธ๋ก์ด ์ข‹์•„์ง€๋‚˜ ํ–ˆ๋”๋‹ˆ ๋‹ค์‹œ ์ œ๋ฏธ๋‚˜์ด๊ธ‰์œผ๋กœ ๋‹ค์šด ใ…‹ใ…‹
Replying to @itslueul
Grok 4.7 should be roughly on par with Opus 5.0, not 5.1. Better in some ways, worse in others. We need to fix multimodal performance. Grok 4.8 will be a noticeable improvement. Grok 4.9 is probably Astra/Fable class. Grok 5 maybe better than anything. We shall see.
86
๊ตญ๋ฏผ๋“ค ์„ธ๊ธˆ ์‚ฅ ๋œฏ๋Š”๊ฑฐ ๊ทธ๋งŒํ•˜๊ณ  ์ข…๊ต ์‹œ์„ค๋“ค ์„ธ๊ธˆ ์ข€ ๊ฑท์–ด๋ผ ์‹œ๋ฐœ ๋ญ” ๋ถ€๋™์‚ฐ ์ทจ๋“์„ธ ์žฌ์‚ฐ์„ธ๋„ ๋ฉด์ œ๊ณ  ๊ธฐ๋ถ€๊ธˆ๋„ ์„ธ๊ธˆํ•˜๋‚˜ ์—†๋ƒ? ์ด๋Ÿฌ๋‹ˆ ์‚ฌ์ด๋น„๋“ค์ด ์™•๊ตญ์„ ๊ฑด์„คํ•˜๊ณ  ๋™๋„ค ๊ตํšŒ ๊ฑด๋ฌผ๋“ค ์ฃ„๋‹ค ์‚๊นŒ๋ป”์ฉ์— ํ•ต์‹ฌ์ง€ ์ž๋ฆฌ ์žก๊ณ  ์žˆ์ง€ ๋‚ด๋ˆˆ์—” ์ด๋Ÿฐ๊ฑด ์‚ฌ์—…์œผ๋กœ ๋ณด์ด์ง€ ์ข…๊ต๋กœ ์•ˆ ๋ณด์ธ๋‹ค ์ด ๋ฐ–์— ์ข…๊ต ์‹œ์„ค์€ ๊ฐ•์ œ ์ˆ˜์šฉ๋„ ์•ˆ๋˜๊ณ  ๊ธฐ๋ถ€๊ธˆ ์ด์šฉํ•ด์„œ ํƒˆ์„ธ ์ฆ์—ฌ๋„ ์กด๋‚˜ ๋งŽ์ด ํ• ๊ฑธ? ๋‚œ ๋ฌด์‹ ๋ก ์ž์ง€๋งŒ, ์ข…๊ต๊ฐ€ ์ธ๊ฐ„์˜ ์ •์„œ์ ์ธ ์•ˆ์ •์— ๊ธ์ •์ ์ธ ํšจ๊ณผ๋ฅผ ์ค„ ์ˆ˜ ์žˆ๋‹ค๊ณ  ์ƒ๊ฐํ•ด์„œ ์ข…๊ต์ธ์ด๋‚˜ ์ข…๊ต ์ž์ฒด๋ฅผ ์‹ซ์–ดํ•˜์ง„ ์•Š๋Š”๋ฐ ์ œ๋ฐœ ํƒˆ์„ธ ์ „์ˆ˜ ์กฐ์‚ฌ ํ•˜๊ณ  ๋ˆ ์ข€ ๊ฑท์ž ์„ธ์ˆ˜ ๊ฑท๊ธฐ ๋”ฑ ์ข‹์€ ๋…ธ๋‹ค์ง€๊ฐ€ ์žˆ์ž–์•„
117
์‹ ์ข… ์ข†๋ณ‘์‹  ๋ฐœ๊ฒฌ ใ…‹ใ…‹ ์˜ค๋ž˜์ „ ์ž‘์„ฑํ•œ ์ด๊ฐ•์ธ ๊ฒŒ์‹œ๋ฌผ ์ธ์šฉํ•˜๋ฉด์„œ ์š•๋ฐ•๋”๋‹ˆ ๊ฒŒ์‹œ๋ฌผ์ด ์ „๋ถ€ ์ด๊ฐ•์ธ ์š•ํ•˜๋Š” ๊ฒŒ์‹œ๋ฌผ ์ง€ํ˜ผ์ž ๋ฐœ์ž‘ํ•˜๊ณ  ์ฐจ๋‹จ๊นŒ์ง€ ํ•จ ใ„ทใ„ท @MIC_DDDD
291
๊ทธ๋ก 4.6 ์ข€ ์ด์ƒํ•œ๊ฑฐ ๊ฐ™์€๋ฐ ๊ธฐ๋ถ„ ํƒ“ ์ธ๊ฐ€? ใ…‹ใ…‹ใ…‹
111
Claw๐Ÿ™ retweeted
187
1,934
513
25,937
886,555
Claw๐Ÿ™ retweeted
My dream game would basically be OpenAI taking the kind of agent/simulation stuff they were experimenting with before modern LLMs and actually turning it into a real game. You start a new world with a population on an island or in a city, and from that point on basically nothing is scripted. Every single person is controlled by their own 5.6 Luna instance, with their own memory, personality, needs, goals, plans, relationships, inventory, money, beliefs and ideas. They have to actually live. Find jobs, earn money, buy food, get housing, start companies, trade, farm, invest, make friends, compete, commit crimes, buy property, build things, organize with other people, whatever they decide makes sense. And the society itself is emergent too. There doesn't have to be a government, currency, police, laws or even private property at the start. The agents can invent those things themselves, agree on rules, vote, create institutions, form political groups, build a justice system, overthrow it all again, or just live in complete anarchy. The game engine handles the actual world and ground truth, physics, resources, money, ownership, construction, combat, inventories, etc., while the Luna agents decide what they want, communicate with each other and make plans. So every new world becomes its own completely unpredictable civilization and story. Someone should release something like this as a game.
173
56
15
1,183
83,528
Institutional Price Action Cheat Sheet 2025 1. Market Structure HH (Higher High) โ†’ Bullish continuation HL (Higher Low) โ†’ Confirmation of strength LH (Lower High) โ†’ Bearish continuation LL (Lower Low) โ†’ Confirmation of weakness 2. Liquidity Zones Equal highs/lows = liquidity pools Stop hunts happen before real moves Institutional orders hide below/above obvious levels 3. Order Blocks Bullish OB โ†’ Last down candle before an impulsive move up Bearish OB โ†’ Last up candle before an impulsive move down Retest of OB = high-probability entry 4. Fair Value Gaps (FVG) Imbalance between buyers & sellers Price gravitates to fill gaps before continuation 5. Institutional Entries Break of structure (BOS) + Retest = safest entry Entry near liquidity sweep = sniper trades Confluence of OB + FVG + BOS = institutional precision 6. Risk Management (The Real Edge) 1โ€“2% per trade, never more Stop hunts are normal โ†’ place stops smartly Patience > entries โ€” institutions wait, so should you Golden Rule 2025: Retail chases moves. Institutions create moves. Trade with the creator, not the chaser.
6
157
2
610
32,510
์†Œ์‹  ๋ฐœ์–ธ : Grok 4.5 ์ข‹์Œ ํด๋กœ๋“œ max 20x ๊ณ„์ • ๋‘๊ฐœ, gpt pro 20x ๊ณ„์ • ํ•œ๊ฐœ ์“ฐ๋‹ค๊ฐ€ Qwen 3.8๋„ ์จ๋ณด๊ณ  ๊ฒฐ๊ตญ ํด๋กœ๋“œ ํ•˜๋‚˜ ๋นผ๊ณ  ๊ทธ๋ก ๋„ฃ์—ˆ๋‹ค ํ•œ๋•Œ ์ œ๋ฏธ๋‚˜์ด๋ž‘ ๋™๊ธ‰์ด๋ผ ์ƒ๊ฐํ–ˆ๋Š”๋ฐ 4.5๋Š” ์ง„์งœ ์ข‹๋„ค
166
Claw๐Ÿ™ retweeted
I asked Codex to pull up some stats and I receive on average one DM or email every 6 or so minutes to ask for a reset. I occasionally do oblige if it comes with really solid feedback or banter.
2,764
223
266
8,750
1,281,808
๊ฒฝ์ฃผ์˜ ๋ฌธํ™”์žฌ๊ฐ€ ์™œ ์‹ ๋ผ๊ฐ€ ์•„๋‹Œ ์กฐ์„ ์˜ ๋””์ž์ธ์œผ๋กœ ๋ณด์ผ๊นŒ? ์‹ค์ œ๋กœ ๋‹น์‹œ ๊ธฐ์ˆ ๋ฌธ์ œ๋กœ ์กฐ์„  ์Šคํƒ€์ผ๋กœ ๋ณต์›๋จ ์‹ ๋ผ์˜ ๋””์ž์ธ์€ ๋ถ‰์€์ƒ‰๊ณผ ํฐ์ƒ‰ ์น ์ด ์ฃผ๋ฅผ ์ด๋ฃจ๋ฉฐ ์กฐ์„  ๋ณด๋‹ค๋„ ํ›จ์”ฌ ํ™”๋ คํ–ˆ๋‹ค๊ณ  ํ•จ
1
130
์™€ ์ด๊ฐ•์ธ ์ŠคํŽ˜์…œ ์œ ๋‹ˆํผ ํŒŒ๋„ค ใ…‹ใ…‹
๐ป๐‘ฆ๐‘’๐‘œ๐‘˜๐‘๐‘–๐‘™ ๐‘ ๐‘ก๐‘ฆ๐‘™๐‘’ โค๏ธ๐Ÿค ๐Ÿค Over The Pitch ๐Ÿ›๏ธ Consรญguela aquรญ โžก loom.ly/8NeHo7U
334
Claw๐Ÿ™ retweeted
I came across a PDF that MIT students compiled, and it's genuinely one of the best quant interview resources I've seen. They call it the "Quant Bible" a single document that reads less like a study sheet and more like an operational playbook. Full version in replies
8
95
2
912
72,367
20๋…„์ „ ์ •์น˜์  ๋ฐ˜์ผ์ •์„œ๊ฐ€ ์‹ฌํ–ˆ๋˜๊ฒŒ ์„ธ๋Œ€๊ฐ€ ๋ฐ”๋€Œ๋‹ˆ ํ™•์‹คํžˆ
์ผ๋ณธ ํ˜ธ๊ฐ๋„ ์—ญ๋Œ€ ์ตœ๊ณ  54.3% ๋‹ค๋“ค ์ผ๋ณธ ์ข‹์•„ํ•˜๊ธด ํ•ด
2
102
๋ผ๋ฉด์€ ์ •๋ง ์™„๋ฒฝํ•œ ๊ฐ€๊ณต์‹ํ’ˆ์ด๋ผ ์ƒ๊ฐํ•จ ๋ผ๋ฉด์— ํŒŒ ๊ณ„๋ž€ ์™ธ์˜ ์žฌ๋ฃŒ๋Š” ์‰ฝ์ง€ ์•Š๋‹ค ํ˜ธํ…”์—์„œ ํŒŒ๋Š” 3๋งŒ์› ๋„˜๋Š” ๋ผ๋ฉด๋„ ๊ทธ๋ƒฅ ๋ผ๋ฉด ๋ง›์ž„ ์—ฌ๋Ÿฌ๊ฐ€์ง€ ํ•ด์‚ฐ๋ฌผ์„ ๋„ฃ๊ณ ๋„ ๋ฐธ๋Ÿฐ์Šค๋ฅผ ์ง€ํ‚จ ์ •๋„์˜ ๊ฐ๋™์ด์ง€ ํŠน๋ณ„ํ•จ์€ ์ „ํ˜€ ์—†์—ˆ๋‹ค ๊ด€๊ด‘์ง€ ํ•ด๋ฌผ๋ผ๋ฉด๋„ ๊ธฐ๋ถ„์œผ๋กœ ๋ง›์žˆ๋Š”๊ฑฐ์ง€ ์ž์ฃผ ๋จน์œผ๋ผ๋ฉด ๋ชป ๋จน์Œ ๊น€๋ฐฅ์ฒœ๊ตญ์—์„œ ๊ณ„๋ž€๋งŒ ๋„ฃ๊ณ  ์ดˆ๊ณ ํ™”๋ ฅ์—์„œ ๋“์ธ ๋ผ๋ฉด์ด ์ œ์ผ ๋ง›์žˆ๋‹ค
1
206
Claw๐Ÿ™ retweeted
Bienvenido al Atleti, Kang In โค๏ธ๐Ÿค Acuerdo con el Paris Saint-Germain para el traspaso del internacional coreano, que firma por nuestro club hasta el 30 de junio de 2031. โžก๏ธ loom.ly/ywNBlFw
447
2,854
786
14,824
1,464,223
X๋ณด๋‹ค ์Šค๋ ˆ๋“œ๊ฐ€ ์•Œ๊ณ ๋ฆฌ์ฆ˜์ด ํ›จ์”ฌ ํŽธ์•ˆํ•จ X๋Š” ๋ญํ•˜๋‚˜ ์žฅ๋ฅด์— ๋ฐ˜์‘ํ•˜๋ฉด ๊ทธ ์žฅ๋ฅด๋งŒ ์ž๊ทน์ ์œผ๋กœ ๋…ธ์ถœํ•ด์„œ ์‚ฌ๋žŒ์„ ํ”ผ๋กœํ•˜๊ฒŒ ๋งŒ๋“ฌ ์ด๊ฑธ ์•Œ๊ณ ๋ฆฌ์ฆ˜ ๋•๋ถ„์— ์‚ฌ๋žŒ๋“ค์ด ๋” ํ™œ๋™์ ์ด๊ฒŒ ๋˜๊ณ  ์žˆ๋‹ค๊ณ  ์ž์œ„ํ•˜๊ฒ ์ง€
1
1
104
OKX๋“ฑ ํ•ด์™ธ๊ฑฐ๋ž˜์†Œ๋“ค์ด ํ”Œ๋ ˆ์ด์Šคํ† ์–ด์—์„œ ์‚ฌ๋ผ์ง„์ด์œ  ๊ตญ๋‚ด๋ฒ• ๊ธฐ์ค€ ๋ฏธ์‹ ๊ณ  ๊ฑฐ๋ž˜์†Œ๋ผ๊ณ  ์—†์• ๊ณ  ์žˆ๊ณ  ๋ฐ”์ด๋‚ธ์Šค ๊ฐ™์€ 1ํ‹ฐ์–ด ๊ฑฐ๋ž˜์†Œ๋„ ์‹ ๊ณ ๋  ๊ฐ€๋Šฅ์„ฑ์€ ๋‚ฎ์Œ ๊ถ๊ทน์ ์ธ ๋ชฉ์ ์€ ํ•œ๊ตญ์ธ๋“ค ์ž๊ธˆ์ด ํ•ด์™ธ ํฌ๋ฆฝํ†  ์„ ๋ฌผ์ชฝ์œผ๋กœ ๋‚˜๊ฐ€์ง€ ๋ชปํ•˜๊ฒŒ ๋ง‰๋Š” ์ทจ์ง€์ž„ ๋‚˜์ค‘๊ฐ€๋ฉด ์ธ์ฆ๋œ ๊ฐœ์ธ์ง€๊ฐ‘ ๊ฑฐ์ณ์„œ ์—…๋น„ํŠธ๋กœ ๋ˆ ๋ฐ›๋Š”๊ฒƒ๋„ ์ž๊ธˆ์†Œ๋ช…์ด ํ•ด์™ธ ๊ฑฐ๋ž˜์†Œ ์ถœ์ฒ˜๋ผ๋ฉด ๊ฑฐ์ ˆ๋  ๊ฐ€๋Šฅ์„ฑ์ด ๋†’์Œ ํ•ด์™ธ ํฌ๋ฆฝํ†  ์„ ๋ฌผ ์ž์ฒด๊ฐ€ ๋ถˆ๋ฒ•์€ ์•„๋‹ˆ์ง€๋งŒ, ๋ฏธ์‹ ๊ณ ๋œ ๊ฑฐ๋ž˜์†Œ ์ถœ์ฒ˜ ์ž๊ธˆ์ด๋ผ๋Š” ํ•‘๊ณ„๋กœ ๋ง‰๊ธฐ ์‹œ์ž‘ํ• ๊ฑฐ๋‹ค ํ•ด์™ธ๊ฑฐ๋ž˜์†Œ ์„ ๋ฌผ ๊ด€๋ จ ์ธํ”Œ๋ฃจ์–ธ์„œ๋“ค์ด ๋ฒŒ์–ด๋“ค์ด๋Š” ์ˆ˜์ˆ˜๋ฃŒ ๊ธˆ์•ก์ด ์–ด๋งˆ์–ด๋งˆ ํ•œ๊ฒƒ ๋•Œ๋ฌธ์— ๋น ๋ฅด๊ฒŒ ์ด๋ ‡๊ฒŒ ๋œ ๋“ฏ ํ•œ๋ฐ, ์‹œ๋Œ€ ์ฐฉ์˜ค์ ์ด๋ฉฐ ๋งค์šฐ ์ค‘๊ตญ์Šค๋Ÿฝ๊ณ  ๋ณ‘์‹ ๊ฐ™์€ ์ •์ฑ…์ž„ ์•„๋งˆ ํ•ด์™ธ ์„ ๋ฌผ 1ํ‹ฐ์–ด ํŠธ๋ ˆ์ด๋”๋“ค์€ ์ด์ œ ๊ตญ์ ์„ ์˜ฎ๊ธฐ๊ฒŒ ๋  ๋“ฏ ์ฃผ์‹๋„ ํ•œ์ฐฝ ์‚ฌ๋žŒ๋“ค์ด ๋ฏธ์žฅ ํˆฌ์žํ•˜๋Š”๊ฒŒ ์œ ํ–‰์ผ๋•Œ ์ง์ ‘์ ์ธ ๊ทœ์ œ ๋ชปํ•˜๋‹ˆ๊นŒ ๋ ˆ๋ฒ„๋Ÿฌ์ง€ ์ณ๋งŒ๋“ค์–ด์„œ ์ง€๊ธˆ ์‚ผ์„ฑ ํ•˜์ด๋‹‰์Šค ์ฐจํŠธ ๋ง์น˜๋Š” ๋ณ‘์‹ ์ง“๋งŒ ํ•จ
3
375
ํ€€ํŠธ ์ˆ˜ํ•™๊ณต์‹๋“ค์ด ์‹ค์ œ๋กœ ์ด ์‚ฌ๋žŒ์ด ๋งํ•˜๋Š” 5๋…„์ฐจ ๋ฐ์ดํ„ฐ๋ฅผ ์œ„์ฃผ๋กœ ์งœ์—ฌ์ง„๊ฒŒ ๋งŽ์Œ ๋‹จ์ˆœํ•˜๊ฒŒ RSI ์–ผ๋งˆ ์ด์ƒ ์ดํ•˜์— ๋งค์ˆ˜ ๋งค๋„ ํ•˜๋Š” ๊ทธ๋Ÿฐ ๋‹จ์ˆœํ•œ ๋กœ์ง์€ ์ˆ˜๋™๋งค๋งค์— ์ ํ•ฉํ•จ
Year 1 as a Trader: โ€ข RSI โ€ข EMA โ€ข MACD โ€ข Supertrend โ€ข Bollinger Bands โ€ข Fibonacci Retracement โ€ข VWAP โ€ข Ichimoku Cloud โ€ข Volume Indicators โ€ข Stochastic Oscillator โ€ข ATR โ€ข ADX โ€ข Elliott Wave โ€ข Moving Average Crossovers 5+ Years as a Trader: โ€ข Price Action โ€ข Market Structure โ€ข Support & Resistance โ€ข Supply & Demand โ€ข Trend Analysis โ€ข Liquidity โ€ข Volume Profile โ€ข Risk Management โ€ข Patience โ€ข Trade Management โ€ข Psychology โ€ข Timing The longer you trade, the less indicators matter and the more you learn to read the market itself.
125
์ž๋™๋งค๋งค์— ์‹คํŒจํ•˜๋Š” ์ด์œ ๋Š” ๋‹จ์ˆœํ•จ ํด๋กœ๋“œ๋ฅผ ๋„ˆ๋ฌด ๋ฏฟ์—ˆ๊ธฐ ๋•Œ๋ฌธ์ž„ ๊ฐœ์ž…์—†์ด AI๊ฐ€ ์•Œ์•„์„œ ๋กœ์ง์„ ๋งŒ๋“ค์–ด์ฃผ๊ธธ ์›ํ•œ๋‹ค๋ฉด ์ตœ์†Œํ•œ AI๋ฅผ ๋งŽ์ด์จ์„œ ์—ญํ• ์ด๋ผ๋„ ๋ถ„๋ฆฌํ•ด์•ผํ•จ
1
1
173
ํ•œ ์ˆ˜ ๋ฐฐ์› ์Šต๋‹ˆ๋‹ค
๋…ผ๋ž€ ์ค‘์ธ ์›Œ์…”์•ก ํ…Œ๋Ÿฌ ์˜์ƒ ์•„๋งˆ๋„ 1์ฐจ๋กœ์—์„œ ์ •์† ์ฃผํ–‰ํ•˜๋Š” ์ฐจ๋Ÿ‰์— ๋นก์นœ 2์ฐจ๋กœ ์ฐจ๋Ÿ‰์ด ์นผ์น˜๊ธฐํ•˜๋ฉด์„œ ์›Œ์…”์•ก์„ ๋ฟŒ๋ฆฐ ๊ฒƒ ๊ฐ™์€๋ฐ... ๊ตญ๋„ 1์ฐจ๋กœ๋Š” ๊ณ ์†๋„๋กœ์™€๋Š” ๋‹ค๋ฅด๊ฒŒ ์ถ”์›”์ฐจ์„ ์ด ์•„๋‹ˆ๊ธด ํ•จ. ๊ตญ๋„์ด๊ธด ํ•˜์ง€๋งŒ 1์ฐจ๋กœ ์ •์† ์ฃผํ–‰์ด ๋” ๋‚˜์˜๋‹ค VS ์นผ์น˜๊ธฐ์— ์›Œ์…”์•ก ํ…Œ๋Ÿฌ๊นŒ์ง€ ํ•˜๋Š” ๋†ˆ์ด ๋” ๋‚˜์˜๋‹ค ์ฐธ๊ณ ๋กœ ์ฃผํ–‰ ์ค‘์— ์œ ๋ฆฌ์ฐฝ ๋‹ฆ์œผ๋ ค๊ณ  ์›Œ์…”์•ก ๋ฟŒ๋ฆฌ๋ฉด ๋’ท์ฐจ๋Ÿ‰๊นŒ์ง€ ์›Œ์…”์•ก์„ ๋งž์„ ์ˆ˜ ์žˆ์Œ. ์–ด์ง€๊ฐ„ํ•˜๋ฉด ์ •์ฐจ ์ค‘์ด๊ฑฐ๋‚˜ ๋’ท์ฐจ๊ฐ€ ์—†์„ ๋•Œ ๋ฟŒ๋ฆฌ๋Š” ๊ฒƒ์„ ์ถ”์ฒœ.
81