Video yükleniyor...

Video Yüklenemedi

Ana Sayfaya Dön

My current controller settings. #BlackOps6 1.60-1.60 0-0 Deadzone 120FOV Dynamic response curve Right stick max threshold 99 Left stick max threshold 75 Hyrbid input

145,063 görüntüleme • 1 yıl önce •via X (Twitter)

10 Yorum

Frankie profil fotoğrafı
Frankie1 yıl önce

forgot to add: Green enemies (easier to see) people I'm partied up with: Dark blue Random players on my team: Purple

Richard 🇺🇸 🎮 profil fotoğrafı
Richard 🇺🇸 🎮1 yıl önce

dont forget to mention "Aim Assist Activated"

blu profil fotoğrafı
blu1 yıl önce

We are still 2boxxing get a life nigga

HB profil fotoğrafı
HB1 yıl önce

Why 99max and not 100? 👊💜

Frankie profil fotoğrafı
Frankie1 yıl önce

It might be placebo but it feels better for some reason

☾ XX profil fotoğrafı
☾ XX1 yıl önce

you forgot to add in unemployment

ap profil fotoğrafı
ap1 yıl önce

wtf is hybrid input

Frankie profil fotoğrafı
Frankie1 yıl önce

Right stick clicked in makes you slide Left stick clicked in and then clicking the right stick will make you dive

𝐋𝐮𝐤𝐞 profil fotoğrafı
𝐋𝐮𝐤𝐞1 yıl önce

Hybrid is king

Frankie profil fotoğrafı
Frankie1 yıl önce

Only thing I would say is, it's a pain in the ass to learn

Benzer Videolar

⚙️Update Quick Video Review: GameSir Connect V1.14.4⚙️ GameSir Connect has just been updated to version V1.14.4. Over the past few updates, we have introduced many practical features to enhance your experience. Here is the video detailing all the highlights of the recent updates. We highly recommend that players visit the Microsoft Store to download the latest version from here: For details on the updates, please refer to the text below. ✅Feature 1 - Professional Tips and Explanation for Controller Settings. You can now hover your mouse over the ⓘ icon to access detailed explanations of controller terminology. These tips have been verified by the GameSir product team to ensure you can quickly, conveniently, and accurately understand the information you need. ✅Feature 2 - Enhanced Stick Customization: Expanded Curve & Calibration Settings Now, you can adjust the stick center point to fix any potential center offset after the controller has already been calibrated. In addition, you can now achieve more precise fine-tuning of the stick input curve by manually entering the coordinates for each point within a 100x100 coordinate system. Please be aware that when Zero Deadzone is enabled, it is normal for the center point to not return to absolute zero or for the data to fluctuate repeatedly. ✅Feature 3 - Support for in-app software updates and halved input latency We've now set up a dedicated server for GameSir Connect software updates. Now, you can update directly within the software. And through optimizations, the end-to-end latency for both the G7 Pro 8K PC and Tarantula 8K PC has now reached 0.25ms with the latest firmware. And only by using the latest software will you be able to receive the latest firmware updates.

GameSir

29,311 görüntüleme • 3 ay önce

I'm using macOS for more than 14 years now, in a single laptop screen setup with lots of windows open, and I'm using macOS spaces everyday to organize my windows. Switching between spaces is easy: I hit CTRL+arrow left/right, it takes no time. What takes time however, is this macOS-native transition animation that takes a bit less than 1 second, for every single space switch. I've checked everywhere in macOS settings, asked friends with the same problem, and there's just no setting to disable this effect (even with Reduced Motion ON in Accessibility settings). This UX choice from Apple alone costs me days of my time every year, as I'm working non-stop on my SaaS business for 10h a day 5-6 days a week (Crisp for those who know about it!). I calculated that with my daily use of macOS spaces, in 6 months I'd perform 100,000 switches, which costs me the same in seconds waiting for the transition to complete. This represents almost 28 hours waiting. In 1 year, I'd have lost 2.4 days just... waiting! The Hack Fortunately, some other open-source projects discovered a hack not so long ago, which allows disabling the transition when switching spaces: it involves sending to the system a "fake" trackpad left/right swipe gesture that's faster than what's expected (faster than what a human can input via the Macbook's trackpad). Above some threshold, macOS just perform the space switch with absolutely no transition. A friend of mine, Yaël 🛸, made a quick prototype in Objective-C, and it magically worked! The Solution So, I asked my friend if we could make it a proper macOS app together, in a beautiful minimalist package, and he said yes! So, we launched Space Rabbit, a tiny 2MB fully native macOS app, that allows you to optimize macOS spaces for productivity! Space Rabbit is a full rewrite of the initial prototype in modern Swift, with a tray menu bar to configure the app, and a settings window, in a macOS app packaging. Space Rabbit can be downloaded right here: We just released V2 of the app last week, fixing a ton of little bugs, especially for dual screen setups. It's actively maintained to this day as it's my daily driver. The app is open-source: and will always be free, as it's helping improve our productivity in our daily Web Dev tasks for our main businesses. We'll be happy to implement more features from user requests, if you find this app useful and use it daily!

Valerian Saliou

222,224 görüntüleme • 2 ay önce

Who is next? This AI Video is absolutely insane Seedance 2.5 on DomoAI official prompt Resolution: 4K, 60fps Style: Ultra-photorealistic Korean action-drama cinematography, realistic live-action performances, fast-paced physical combat, natural daylight, dynamic handheld camera movement. SETTING A modern South Korean high school during daytime. A long, straight corridor with large rectangular windows along the left wall, beige classroom walls, wooden-framed glass classroom doors along the right, polished light-gray floors reflecting sunlight, white fluorescent ceiling lights, green exit signs, and red fire extinguishers positioned against the walls. Bright natural afternoon sunlight enters through the windows, creating strong rectangular patches of light across the corridor floor. Green trees and distant apartment buildings are visible outside. The entire action takes place within this same continuous school corridor. CHARACTERS FEMALE LEAD Grok-image-b16c1304-9c31-4c04-842d-d50f1a719e65 (1) She wears an elegant pastel-pink cropped school-uniform-style blazer with structured shoulders, long sleeves, a tailored waist, notch lapels, front buttons, and a small embroidered school crest on the chest pocket. Underneath is a crisp white collared shirt with a large pink-and-burgundy plaid ribbon bow tied neatly around the collar. She wears a high-waisted, short pink plaid pleated skirt with a soft A-line silhouette, featuring subtle gray and burgundy check patterns. Her outfit is completed with white knee-high socks featuring two delicate pink horizontal stripes near the top and chunky white platform sneakers. Her movements are extremely quick, agile, athletic, and physically convincing. Her hair flies naturally during running, turning, and fighting. MALE LEAD: A handsome Korean man, 22, with slightly messy short purple hair, fair skin, and a slim athletic physique. He wears a dark charcoal school blazer with matching green school badge, black V-neck sweater, white shirt, burgundy-red tie, and black school trousers. He has a surprised, concerned expression during the final confrontation. MALE ATTACKERS: Four Korean men, ages 21–24, wearing slightly disheveled dark school uniforms, purple trousers, and sneakers. Some have loosened ties or open blazers. Their expressions are aggressive, frantic, and determined. One carries a short wooden stick. All characters must maintain consistent faces, hairstyles, uniforms, and body proportions throughout. SHOT-BY-SHOT ACTION SHOT 1 — 0:00–0:02 | THE THREAT Medium rear-facing shot inside a classroom doorway. The female lead stands with her back toward the camera, pushing open a large beige wooden classroom door with both hands. Several angry male students in dark uniforms stand clustered inside the doorway, blocking her path. One male student stands prominently in the center, staring at her. She immediately realizes she is surrounded. Camera: Static medium shot with a slight handheld tremor. Natural daylight and realistic shadows. SHOT 2 — 0:02–0:04 | SUDDEN ESCAPE Hard cut to the long school corridor. The female lead bursts into the hallway and runs toward the camera at full speed. Her black hair whips across her shoulders, her blazer flutters, and her plaid skirt moves naturally with her rapid footsteps. She glances sideways while extending one arm for balance as she accelerates. Camera: Fast backward tracking shot, slightly low angle, 35mm lens. Background motion blur while her face remains relatively sharp. SHOT 3 — 0:04–0:06 | THE PURSUIT Wide tracking shot looking down the hallway. Four male attackers sprint after her, shouting aggressively. The nearest attacker leans forward while running, almost losing his balance. Another follows with his mouth wide open, while a third carries a wooden stick. Their shoes strike the polished floor loudly. Camera: Handheld backward tracking, rapid movement, subtle camera shake, energetic action-film pacing. SHOT 4 — 0:06–0:10 | EXPLOSIVE CORRIDOR FIGHT Cut to a wide side-angle shot near the large corridor windows. The female lead abruptly stops running, plants her feet wide apart, and turns to confront the approaching attackers. The first male attacker rushes toward her. She quickly evades his charge, causing him to stumble and crash onto the floor. A second attacker charges with a wooden stick raised above his shoulder. She twists her torso, shifts her weight, and dodges his swinging attack. Another attacker loses his footing and slides across the polished floor behind them. The action is fast, aggressive, and physically grounded, with convincing momentum and clear spatial continuity. Camera: Wide 24mm lens, dynamic handheld movement, rapid lateral tracking. Keep full bodies visible during major movements. SHOT 5 — 0:10–0:12 | THE KICK Extreme close-up of the female lead delivering a powerful forward kick. Her white sneaker drives into the torso of an approaching attacker, compressing the dark fabric of his uniform at the point of contact. His upper body recoils backward from the force. Cut immediately to a close-up of her face. She stands still, breathing hard, her expression cold and focused, with strands of black hair falling around her cheeks. Camera: Fast impact close-up followed by a steady 85mm portrait shot. Shallow depth of field, realistic fabric deformation, and natural skin texture. SHOT 6 — 0:12–0:15 | THE WOODEN STICK Wide corridor shot. A male attacker swings a short wooden stick horizontally toward the female lead. She rotates sharply, her hair whipping outward, and narrowly avoids the strike. The wooden stick passes across the foreground, creating a dramatic motion blur. She completes her turn and regains her balance near the windows. Camera: Fast lateral tracking shot with a brief whip-pan following her rotation. Natural-speed combat without exaggerated slow motion. SHOT 7 — 0:15–0:18 | UNEXPECTED ENCOUNTER Cut to a medium frontal shot of the male lead standing in the corridor. He is motionless, staring ahead with widened eyes and slightly parted lips. His expression shows surprise and disbelief at the chaotic fight unfolding before him. Cut to an over-the-shoulder view from behind the female lead. One of the attackers suddenly turns and runs away down the long corridor, disappearing into the distance. The hallway becomes momentarily quiet. Camera: Steady medium close-up followed by a long-lens hallway perspective emphasizing depth and distance. SHOT 8 — 0:18–0:24 | FINAL CONFRONTATION Cut to a medium close-up of the female lead standing in the corridor. She faces the male lead, breathing heavily after the fight. Her black hair is slightly disheveled, her uniform remains intact, and her expression shifts from fierce determination to cautious surprise. She stares directly toward him. The bright school windows behind her create a soft natural rim light around her hair. She remains nearly motionless as the camera slowly pushes closer. A short Korean male voice is heard from off-screen: "들어와." (Come in.) The female lead reacts with a subtle change in expression, her eyes widening slightly. Hold on her face for the final moment. End with a clean cinematic cut. CINEMATOGRAPHY Photorealistic Korean action-drama production quality. 24mm wide-angle lens for full-body fight choreography. 35mm handheld tracking for running and chase sequences. 85mm portrait lens for emotional close-ups. Fast, sharp editing with continuous action geography. Bright natural school daylight, slightly cool neutral color grading. Realistic skin texture, visible pores, natural facial imperfections. Realistic motion blur, weight, gravity, inertia, and physical contact. Natural handheld camera shake without excessive instability. Convincing stunt choreography with clear cause-and-effect movements. SOUND DESIGN No background music, no voiceover, no narration. Only natural production audio: rapid footsteps, shoes squeaking against the polished floor, heavy breathing, distant male shouting, fabric rustling, wooden stick swinging through the air, body impacts, and natural corridor reverberation. Use natural Korean dialogue where indicated. STRICT CONSISTENCY RULES Maintain the same female lead and male lead throughout. Never change their faces, hairstyles, clothing, or proportions. Maintain one consistent Korean school corridor with the windows on the same side and classroom doors on the opposite side. Every kick, dodge, stumble, and fall must follow realistic human biomechanics. Avoid floating bodies, unrealistic flips, teleporting characters, duplicated attackers, disappearing props, broken anatomy, inconsistent spatial positions, and artificial-looking CGI movements. No supernatural abilities, glowing effects, stylized animation, exaggerated wirework, or fantasy elements. FINAL RESULT: A fast-paced, believable, live-action Korean school action sequence featuring a sudden escape, intense hallway pursuit, aggressive close-quarters combat, a dramatic encounter with a male student, and a tense final close-up. #DomoAI #DomoAICPP

Sharon Riley

20,041 görüntüleme • 10 gün önce

My BIGGEST Takeaways For DEFENSE In #NBA2K27 So Far…🔒 After Playing And Testing Defense, There Are A LOT Of Things I’ve Started To Notice About What You ACTUALLY Need On Your Builds This Year… And You Have To STOP Playing Defense Like This Is NBA 2K26 Here’s What I’ve Learned So Far: - BLOCK - 84 - 93 Is The SWEET SPOT For Lockdowns And Bigs, I Think 84 - 93 Block Is Going To Be ALL You Need Throughout The Year, Assuming We Don’t Get Any Drastic Gameplay Changes For Primary Ball Handlers, 60 Block Is Actually Enough To Give You A Chance At Getting Stops. Anything Lower Than That Don’t Expect To Get Many Blocks At All Overall BLOCKS ARE OVERPOWERED RIGHT NOW - ON-BALL DEFENSE IS AMAZING On-Ball Defense Feels YEARS Ahead Of What We Had In NBA 2K26. But The Biggest Change Is Learning That You CAN’T Defend The Same Way We Have In Previous 2Ks. Your Positioning And What Defensive Input You Use At The Right Time Matters A LOT More - STOP OVERUSING RIGHT STICK DEFENSE Using Your Right Stick For Hands Up Defense And Shot Contests Works REALLY Well When You’re Trying To Contest A Shooter At The EXACT Moment They Pull Up. But If You Start Using It Too Early It Slows You Down A LOT And Can Give The Ball Handler An EASY Blow By Or Drive. The Right Stick Can Either HELP Your Defense A LOT Or HURT You Depending On WHEN You Use It - TURBO + INTENSE DEFENSE IS THE KEY Start Using Turbo + Intense Defense (RT/LT Or R2/L2) As A Combination. Learning When To Go Back And Forth Between The Two Makes It MUCH Easier To Stay In Front Of Your Matchup, Cut Off Driving Lanes And Get Into Better Positioning. Most Importantly, It Helps You Stay In Position And Preserve Your Energy For Blocks - PERIMETER DEFENSE + SPEED + AGILITY Perimeter Defense Works Hand In Hand With Your Speed And Agility In. Right Now, I Think: 82 - 92 Perimeter Defense, 85 - 90 Speed + Agility Is MORE Than Enough To Feel Great Defensively In The Current State Of The Game. You Don’t Need To Throw 99 Into Every Defensive Attribute To Stay In Front Of People. - STEALS FEEL OVERRATED High Steal Builds CAN Get Some Really Easy Pluck Steals Right Now… But I’ve Also Noticed A Lot Of Those Builds Sacrifice So Much To Get Their Steal That They Become A HUGE Liability Once Someone Gets Into The Paint. Unless Stealing The Ball Is A Major Part Of Your Role, I Wouldn’t Sacrifice Your Entire Defense Just To Get A 99 Steal Rating. - INTERIOR DEFENSE MATTERS A LOT I Think 82 - 89 Interior Defense Is Going To Be The SWEET SPOT. You Really Only Need A Few Important Badges, But DO NOT Let Anybody Convince You That Interior Defense Doesn’t Matter This Year. Because If You Ignore It You’re Going To Be FREE POINTS IN THE PAINT Even If You Feel Like You’re Timing Your Contests Correctly Defense In #NBA2K27 Feels REALLY GOOD Right Now

R1CH SOSA🔋

50,886 görüntüleme • 1 ay önce

Trading without AI in 2026? You're done. Purely on the math, before anyone even out-trades you. You are not up against other people anymore. 75% of US stock volume runs through machines. Jane Street put $6 billion and 4,032 GPUs behind agents that write trading strategies, test them and kill them all night. Man Group's AI turns out hundreds of new signals a week where its human desk used to manage about 20 a quarter. You're on the other side of that with a chart and a feeling. 89% of retail traders lost money last year. The gap was never brains. Testing one idea properly costs a desk about $87,500 in salary time or massive computing power for AI. You don't have any of those things. That part is over. You have your own AI trading agent now, built by a team backed by Entrée Capital and hedge fund managers, and trying it costs nothing: No code, no desk, no excuses left. What it does with one English sentence: > compiles your idea into entry, exit, sizing and risk rules > backtests 5 years of data in about 12 seconds, 1-minute bars in 47 > scores it 0 to 100 on Sharpe, Sortino, profit factor and max drawdown > kills anything with a half-life under 10 days no matter how pretty the curve looks > runs walk-forward and Monte Carlo on whatever survives, then years of data it has never seen > puts it live on your exchange in about 90 seconds and keeps running while you sleep Same machine the funds run, minus the hiring, the salaries and the three-week wait. 4,800+ traders are already inside. Save this and put one idea through it tonight after read full breakdown below. You'll know in 12 seconds what three weeks of hoping never told you.

cvxv666

38,982 görüntüleme • 2 ay önce

My app just hit $7.8k/m in revenue. This is how I'm running ads for mobile apps. To hit 10k/m before I quit my job & fly to Vietnam 🇻🇳 Part One: Apple Ads (First $1k/m). • Repeated this a lot so to summarise: 1. $100 free ad credits, ez mode. 2. Run strict keywords on competitor names. 3. Split into tiers for countries & scale the max bid. 4. Use RevenueCat or Superwall for attribution. 5. Give ad spend & transactions to claude: find winners. 6. Measure the onboarding: tighten funnel. 7. Daily budget can be low at first, like $20. 8. The max bid is the important variable here. 9. This will get you to $1k/m. Part Two: TikSl0p Ads ($1k to $10k) • This is much harder to make profit. • But when you do, it's exponential. • Read Open Book's guide on this first. 1. Make ~6 formats, ~12 creatives. 2. Leave campaign targeting super broad. 3. I do $70/day of campaign budget spend. 4. No demographic or interest settings. 5. Let the algorithm find your perfect user. 6. Give each creative $50 of ad spend. 7. After $50, if (close to) profitable then scale. 8. If no profit, then stop the posts. 9. Use realised LTV to begin with. 10. Simple: revenue / paid customers. 11. 400 IQ chad: aim for profit on initial transaction. 12. Ask Rico for unrealised LTV. Can be hard to calculate. 13. Expect to burn a few hundred $$ to find profit. Notes: If you none of your creatives are profitable, then go back to the drawing board and try again. You're taking $50 gambles each time. When you find a winner, it'll repay the initial investments 100x! This is a game of pain tolerance. You need to be able to stomach losing hundreds of $$$ without seeing any returns. Part Three: Scaling. 1. If you're consistently getting > 2x ROAS. 2. Up your campaign budget by 15%. 3. Do NOT increase more than 30%. 4. Otherwise I will be very very cross with you. 5. You'll kill performance. 6. Performance can be volatile. • Measure ROAS over weeks, not days Notes: When you've found a winning creative, create multiple variations of it. We're talking 100 different kinds. I'm being lazy and need to do this more!! Part Four: Experimental Ad Spend ($10k +). • This is the next level: I've gone 0-1. • I've not seen much sauce on X about scaling past this. • So, this is where I'm at now. I'll share it with all of you. My current position is: 1. Trying out other platforms: Meta, IG, etc 2. Continue reinvesting the profits from previous ads. 3. I'm trying multiple new avenues right now: • Paid influencers, UGC, slideshowmaxxing, etc. Keep pushing, we're all gonna win.

Frederick James

17,553 görüntüleme • 2 ay önce

The current physical activity guidelines undervalue vigorous activity. Vigorous activity may be 4–9× more potent than moderate activity for reducing all-cause mortality, cardiovascular disease, diabetes, and cancer risk. The exercise guidelines assume a 2:1 ratio between moderate- and vigorous-intensity activity: two minutes of moderate activity equals one minute of vigorous. That's why the recommendations are 150–300 minutes of moderate or 75–150 minutes of vigorous activity each week. But new data suggest that ratio is wrong. In this brand-new journal-club episode, Brady Holmer (Brady Holmer) and I unpack a groundbreaking study that should change how we think about activity for disease prevention. The research identified that 1 minute of vigorous activity is roughly equivalent to 4–9 minutes of moderate activity, and 53–94 minutes of light activity, for disease risk reduction. It also shows a clear dose-response for vigorous activity that’s much weaker for moderate activity and barely detectable for light activity. We also do a deep dive into why vigorous activity is so powerful, the underlying mechanisms, and discuss practical takeaways, including how even very brief bouts of vigorous movement (think “exercise snacks”) can produce meaningful health benefits. Timestamps are below. You can find links to the episode on YouTube, Apple Podcasts, and Spotify in the next post. Enjoy! Timestamps: 0:00 - Introduction 2:14 - The 1:2 rule for exercise 6:28 - What counts as vigorous? 8:48 - Where exercise guidelines fail 9:32 - Inside the wearable-based study design 15:24 - Vigorous activity—easier than you think? 18:01 - Avoiding healthy user bias 19:12 - A better way to measure exercise 20:58 - Is vigorous 4–10x better? 25:08 - One vigorous min vs. one-hour walk 27:15 - Are light activity's benefits capped? 29:03 - Is vigorous 5x better for your heart? 30:12 - Does zone 2 count as vigorous? 31:24 - Dose-response comparison 32:35 - Vigorous exercise & younger arteries 38:29 - Why aging hearts need intensity 41:22 - Can intensity preserve VO₂ max? 42:40 - Moderate exercise & VO₂ max limits 44:34 - Is vigorous 10x better for diabetes? 51:01 - Why intensity boosts mitochondria 56:11 - Does intense exercise kill tumor cells? 1:02:28 - Hormonal benefits 1:03:19 - Preventing falls with intensity 1:07:49 - Fighting inflammation 1:09:42 - High-intensity training & brain aging 1:11:14 - The 2:1 ratio is out the door 1:13:03 - Could vigorous exercise become a pill? 1:14:21 - Short bursts for longer life 1:18:28 - Can short bouts match full workouts? 1:22:39 - Do wearables undervalue vigorous bursts? 1:25:19 - Can micro-workouts replace the gym? 1:30:23 - Updating exercise guidelines 1:41:48 - Is light activity useless? 1:44:17 - Is vigorous exercise safe for seniors? 1:48:41 - Is HIIT harmful to female hormones? 1:54:18 - Balancing intensity & recovery (80/20 rule) 1:56:43 - Brady’s exercise routine 2:00:30 - Vigorous activity & kids’ brainpower 2:03:27 - Are we undervaluing vigorous exercise? 2:05:16 - Why chasing steps doesn't work

Dr. Rhonda Patrick

306,762 görüntüleme • 10 ay önce

An Anthropic researcher sat down next to me at a hackathon last week. Claude Opus 4.7 was running 4 agents on my laptop. Live. No manual input. She looked at the terminal and said: "What is this?" I showed her. 4 agents. 678 trades. 81% win rate. $16,200 last 30 days. She worked on the evals team. She'd never seen Claude pointed at 88 million on-chain trades. The setup is 3 public repos. All free. -> 88 million Polymarket trades. Every wallet. Every entry. Every exit. Every resolution. -> the framework that bridges Claude Opus 4.7 directly to live markets. Order placement, position tracking, exit timing. -> real-time WebSocket order book. Depth on both sides. No polling, no lag. Four agents. One loop. Agent 1 identifies which wallets win consistently across 88 million trades. Agent 2 reverse-engineers their entry timing. Agent 3 monitors order book volume spikes. Agent 4 sizes positions using Kelly. No overbet. Drawdown capped at 1.4% over 678 trades. 85% of windows get killed. No trade. The bot only enters when 3 signals align: -> Elite wallet consensus pointing the same direction. -> Price divergence with Binance and Coinbase both agreeing. -> Order book imbalance confirming the bias. Single-source price data was 57% accurate. All three together: 81%. Exit before resolution. Always. Losers hold to 0 or $1. The agents copy their exits. The agents don't gamble on that. My stack: Claude Opus 4.7 at $19/mo, VPS Hetzner at $4.99/mo, Everything else free. Total stats: $23.99/month. 30 days: 678 trades, 81% win rate, net +$16,200, max drawdown -1.2%, avg hold 4h 12m. She asked if Anthropic could test this internally. "We run Claude on benchmarks and evals. Nobody pointed it at a live market dataset with 88 million rows." Claude Opus 4.7 didn't need a system prompt. It read the wallet index, understood the signal structure, and wrote the combiner logic in one pass. The people who built the model hadn't thought to point it at this data. I had. Copy the live trades: -> all 4 agents run 24/7. The window is open right now. Save this, follow me and comment OPUS. I will send the guide to you.

slash1s

46,415 görüntüleme • 6 ay önce

STOP PAYING FOR MOTION DESIGN SOFTWARE opus 5.5 one shot every frame of this in code, no after effects im open sourcing the whole template for free plug in your product and it recreates this steal the prompt ↓ Ask me for: the one question a user asks my product and its one-line answer, two album covers, a royalty-free song around 110 BPM with a quiet intro and a hard drop (Mixkit, free for commercial use), and my logo for the ending. If I skip any, use the defaults: "is this all code?" / "yes. every frame.", two atmospheric photos, Mixkit "Head Bang". A 22 second square loop, 1440x1440 at 60fps, 2D only, one continuous take. Nothing ever cuts: every scene grows out of the last, and every gesture physically carries the camera into the next world. Every morph glides for 0.8s on cubic-bezier(.45,0,.15,1) and the camera zoom rides the same curve, so shapes change so smoothly you barely notice. Content dissolves in the back half of each glide. One gesture per scene. Look: a warm light stage (#f3f3f0 to #e3e3de), black UI, the app world dark (#121212) tinted from the current album cover, one green accent (#1ed760). Geist for UI, the real macOS pointer, soft layered shadows that grow with the shape's height. Banned: hard cuts, bouncy springs, particles, glows, zooming in and straight back out, holds longer than 1s, anything that looks like a template. Beats are the song's, and the press lands on its drop. Generate: the cursor clicks, the pill glides into a spinner, then a check. Chat: the check splits into three typing dots (goo). The left dot flies up into the question bubble and types it, the right one flies down into the reply. Both contract back into dots while gliding in and pour into the middle one. Play: the middle dot bends into a play triangle, point by point. The cursor presses it on the drop: a white button pops in under the click and the dark app floods out around it, so the leftover white becomes the play button. The camera pulls back to reveal the full now-playing screen: big cover, title, heart (liked, fills green), progress, controls, volume. Swipe: the cursor drags the cover sideways, the next track slides in, and the whole app retints from the new cover. Drag: the camera glides to the volume. The cursor drags it to max and keeps pulling: the bar rubber-bands and drags the camera sideways out of the app, tracking the knob across a dark dotted canvas. Chart: the line lands and bends up into a green chart while the number counts up. The volume knob becomes its last point. Dive: the cursor clicks that point and the camera dives into it until the light stage fills the frame. Ending: an "Ask [my product]" button grows out of the light, thins into a line, and the line fans out into my logo (default: the Claude spark, six copies rotating into twelve rays in Claude orange). It folds back into the line, which thickens into Generate. Last frame = first frame. 1. One HTML canvas. Everything is a pure function of time inside seek(t): no timers, no state between frames. 2. A camera with position and log-space zoom. Scenes live side by side in one world, so a pan really travels. 3. Goo: blur + alpha threshold on an offscreen layer for the split and the merges. 4. Floods are circles that grow from whatever caused them to the farthest corner in about 0.4s, and the colour switch happens underneath them. 5. Sound: before the press the song plays through a low-pass filter, like it's in the next room, and opens up to full on the drop. A downloaded Mixkit SFX for every event, each placed by its measured peak. Loudnorm to -14 LUFS. 6. Render with Playwright at 60fps with 4 motion-blur subframes, 12 on the fast pan. Then scan every frame for single-frame jumps and fix each one. A handoff between two drawings only works once the first one has fully landed, or it pops. The cursor must stay on whatever it drags, even while it stretches. A counter or tooltip that changes every frame smears under motion blur, so swap whole values. Ask me for the inputs, then show me the beat map and 6 stills (chat, play, app, drag, chart, logo) before you render.

zero

47,922 görüntüleme • 21 gün önce

opus 5.5 is f*cking insane at motion design every frame is code with 0 external tools im open sourcing the template for all these morph transitions steal it ↓ Ask me for: my product's name (one lowercase word, it becomes a black wordmark with a period), the one job a user hands it before bed, the three results it reports by morning, two photos of the same place at night and by day (aligned, so a crossfade reads as a real sunrise), one sunny daytime photo for outside the logo, and a song with a big hit near its start. If I skip any, use the defaults: "nightshift", "book 10 sales calls this week", 200 leads found / 200 emails sent / 10 calls booked. A 24 second square product launch video, 1440x1440 at 60fps, 2D only, one continuous take that loops. The whole video happens inside the period of the wordmark. Outside it's daytime (the sunny photo behind a black wordmark). Inside it's the night photo under iOS-style glass panels (the blurred photo behind each panel, a tint, a rim light). Nothing cuts: every scene grows out of the last, shapes morph, dots fly, colours change under floods. Every morph glides 0.8s on cubic-bezier(.45,0,.15,1) and the camera zoom rides the same curve. Geist, the real macOS pointer, soft shadows that grow with each shape's height. Banned: flat or gradient backgrounds made in code, camera zooms into tiny points, hard cuts, crossfades between scenes, bouncy springs, particles, glows, holds longer than 1s, cropped type. Beats are the song's. The song enters on its big hit as the first pill appears. Open: the letters slide into the period and vanish into it while it glides to the middle. The period opens like an iris onto the night inside it, and "you log off at 6", then "[product] doesn't", roll through one mask line over the night photo. Ask: a glass "Ask [product]" pill springs in. The cursor clicks it, it glides into an input and the job types out. Find: send shrinks the input to a dot that ripples into a 23x15 grid. A scan line sweeps it, 200 dots light up and a counter counts them. Write: the lit dots pour into a glass email composer. A second, purple cursor labelled "[product]" types a personal opener and presses Send, and the counter runs to 200/200. Night: the composer becomes a glass orb with a progress ring. A clock chip swaps 11:42 PM, 3:17 AM, 5:02 AM, 6:59 AM while the sky relights from night to dawn and "while you sleep" rolls in. Results (beat 20): the orb turns black and goo-splits into the three result pills. The checks pop green. Morning: the pills merge into a Dynamic Island that grows into a phone lock screen: the day photo, the date, a big 7:00, the bezel growing out of the screen edge. Three glass notifications stack in, and the cursor taps the first, which opens like iOS. Proof: a week calendar (10 meetings pop in, a hover shows one's time), then a pipeline chart ($48,200 counts up, the line draws, a hover follows it). A black flood from the line's end becomes "Start your night shift →" and the cursor clicks it. Out: the button shrinks into a dot, the whole scene closes down into it like an iris, and the letters slide back out of the period. The tagline rises. Last frame = first frame. 1. One HTML canvas. Everything is a pure function of time inside seek(t): no timers, no state between frames. 2. One element morphs through every state (size, radius, colour, material) from a table of beat times, so retiming is one edit. 3. Render the inside world into an offscreen canvas and reveal it through the period with a circular mask. Never zoom the camera into the dot. 4. Glass: a pre-blurred copy of the photo clipped to the shape, then a tint and a gradient rim. 5. Goo: blur + alpha threshold on an offscreen layer for the split and the merge. 6. Sound: a downloaded SFX for every event, each placed by its measured peak. The song enters on its hit. Loudnorm to -14 LUFS. 7. Render with Playwright at 60fps with 4 motion-blur subframes. Scan every frame for single-frame jumps and fix each one. A camera zoom into an off-centre dot swings sideways and spikes, so open the dot instead. Ease a closing iris by area, not radius, or its first frames jump. A helper that sets globalAlpha must restore the caller's, or fading cards pop. Counters read the frame's time, not the subframe's, or the digits smear. Ask me for the inputs, then show me the beat map and 6 stills (open, find, write, results, phone, out) before you render.

zero

12,628 görüntüleme • 17 gün önce

If you watch this ~50 minute screen recording closely (yeah, I know, it's long; there are also some times when my computer was very slow and laggy, just skip past that part. And at one point I had to run and get my 9-month-old a new bottle and left it on a boring screen, sorry!), I believe you can see real signs of the kind of runaway, recursive AI self-improvement that people have been warning of for a while (Mr. Kurzweil most notably and prophetically). Why do I say that? What's different now? Well, there's a reason my set of agent coding tooling is called the Flywheel. These tools all mutually self-reinforce each other. And they all flow directly into my ntm tool (short for "named_tmux_manager"), which acts as a sort of integration point and nerve center for the tools (this is becoming more true by the minute as I'm now seriously working on ntm). Now, ntm was something I started making to automate some aspects of my workflow, but it was the kind of thing where, until it was perfect, it sort of just slowed me down. So I didn't actually use it even though I kept working on it and trying to improve it, and suggested to users that they try it in my tutorials. Well anyway, I finally got around to "dogfooding" ntm last night, and now it's going to get very dramatically better at an alarming rate. Some of that is from applying my "idea wizard" prompt to generate more useful features and building that stuff out and addressing obvious pain points I encountered during my newfound usage of the tool. But a lot comes from my realization that, once again, ntm's true utility is not as a tool for ME, but for an agent. That is, ntm lets one instance of Claude Code or Codex act as, well, me, do the things that I had been doing manually. Do I wish I had started using ntm earlier? No, for two big reasons: 1) Doing it manually helped me build up my intuition massively, which directly led me down the path of creating useful prompt strategies and workflows; these often began as ad-hoc prompts that I realized could be generalized and made more versatile/universal. Lesson: don't prematurely automate until you have an intimate, intuitive feel for your "core value-add loop." Otherwise you'll have a fully automated system quickly that efficiently and automatically does a stupid or otherwise sub-optimal thing. 2) My eyes have been opened to the beauty and power of Skills. I'm not talking about your garden-variety skills that are just a simple markdown file. I'm talking about true tour-de-force directories of perfectly structured and organized files that are filled with good information, insights, workflows, etc., but presented in a way that is highly optimized for consumption by AI agents, with extreme attention paid to things like perfect progressive disclosure, token density, agent-ergonomics, agent-intuitiveness, etc. And also Skills that go way beyond markdown files, with full integration into Claude Code where it makes sense via hooks, sub-agents, and even Python scripts. These kinds of skills are a qualitative difference in expressive power and usefulness and a total game changer. They are also effectively composable, creating almost an algebra of skills that let you use them together in powerful ways. I'm working on a subscription service website and CLI tool now to share what I've learned here most effectively, stay tuned for that in the coming days. Anyway, I now know what to make and how to make it. So, getting back to that screen recording, what does it show that makes me claim recursive self-improvement is here? If you keep your eye on the upper left tmux pane, that's the "controller" agent. It is using ntm to control all the other panes which are also running Claude Code (but ntm fully supports other agent types like Codex and Gemini-CLI, and it's trivially easy to mix and match them if you wanted to have, say, 8 CCs and 6 Codexes for writing the code and 3 Gemini-CLIs for reviewing code.) Now, there's nothing that crazy about this much so far. But where it starts to get very cool is that as the session continues and we encounter real-world problems, things like my ridiculously overloaded computer that keeps hanging for long periods, Claude Code instances that crash and get into a frozen, unresponsive state, it can learn from that. And you can see it using my skill writing skill to refine its ntm vibe coding skill in real time. And then take that skill and refine it to be more intuitive for itself. Or use my cass tool skill to search all the session histories to look for problems that came up and strategize how to solve them. The most useful part was when, towards the end of the session, I told it to reflect on all the things we had done and problems we encountered. One way it can usefully leverage those reflections is by improving its ntm vibe coding skill to make it cover more edge cases and exigencies. But the other, more fundamental, way is for it to conceive of and design the optimal new features and functionality for ntm itself so that the tool embodies those lessons in a first-class way. This offloads cognition from its brain onto its tooling, just like how a person can lean on spellcheck or a calculator. It codifies correct, effective reasoning at the tool level, where it's more reliable and robust and repeatable. And btw, did you notice what code base it was working on the whole time? It was none other than ntm itself! So as it worked on its own tool, it had reflections and ideas about how to further improve the tool. Now, it could have just as easily gotten those insights and ideas while using ntm to work on a different project, but the fact that it was working on itself is almost gloriously meta and recursive. So by the end, after learning from tending to a big group of agent workers (btw, I have previously emphasized doing everything in a really distributed/decentralized way, where each fungible agent gets identical marching orders that tell it to use my bv tool to find the optimal bead to work on. This does work very well, but occasionally results in some contention and overlap from thundering herd, or at least wastes time/tokens/communication in avoiding that before the agents waste time duplicating work. But in this new ntm-oriented workflow, I was able to have the controller agent in the upper left use bv itself and then optimally parcel out the instructions to each agent so that we could know for sure that there's no overlap), I ended up with a ton of new beads for new features, which I had it optimize and polish a few times. Now I can swap to a new Claude Max account and have the swarm implement all those new features! It should only take a couple passes like the one shown in the screen recording to get everything implemented. Then we can rinse and repeat, having the agent read through the full session histories of each agent and its experience from its own session in sending ntm commands and seeing how they worked out in practice, to come up with the next batch of changes to both its ntm vibe coding skill AND to the ntm tool itself. Do you see how rapidly this turns into Skynet? My mistake earlier was in focusing on making myself a "faster horse" as Henry Ford used to joke about customers wanting before he showed them what they should really want (a Model T). That is, something that would make my experience nicer while doing this agent swarm based development workflow. But the obvious lesson is that you should make all your tooling agent-first because the agents are just better at this stuff. You can still watch, and of course I did add a ridiculous number of very nice human-centric features to ntm that you'll be seeing in the next day or two, but those are really kind of "for fun" to make us humans feel better about the process. All the real value-add is happening "by agents, for agents." PS: Towards the end, you can see me switch to my Mac and tell Claude to improve the skill that I made earlier today for taking the mkv screen recording files from OBS Studio and muxing them into MP4 files for sharing, while downloading songs from YouTube to serve as the background music. I made it so it can also grab the thumbnails and generate little song credit cards that show up in the lower right corner. This worked perfectly the first time! I'll include some screenshots in a response post showing how that worked, but it was awesome to witness. Skills are POWERFUL. I'll also post a link to this video on YouTube if you prefer to watch it there.

Jeffrey Emanuel

25,483 görüntüleme • 9 ay önce

Surprising progress from ~3 months of teaching the 8yo to do "keepy uppy" with a soccer ball. Some lessons learned... First, this skill is embedded in a larger goal that he cares about. He wants to be a better footballer (I hate calling it "soccer", which comes from a contraction of "association" 🤮). This means there is some baseline motivation. This helps you get through the times it sucks to practice. And by the way... It's ok if it sucks (just not too much). What you don't see here is him failing 1,000 times and rage quitting. Probably 1/3 of the sessions involve tears. But you get the biggest gains when you're in your "zone of proximal development" -- the place where it's not too hard, not too easy. It took a lot of cajoling to get him through the early sessions when there wasn't much progress yet. It has to be his choice. When he gets upset, I don't force him. I just ask him if he wants to quit. He never does. I think if I forced him to stick with it, he would be less motivated. By asking him to step outside his momentary frustration and make the decision to quit or not, I think I'm helping him have the space to think about how this connects to his larger goal. But I don't really know. Human motivation is mysterious. Long consistency + intensity, short duration. We do 15 minutes a day, every day. First thing in the morning. I believe duration of training is overrated. In school for example, I think we could get more learning in 30 intense minutes per day than we get in 6 hours a day plus homework now. It takes more effort to do 15 minutes of keepy uppy than it takes to do 2 hours of team practice. And I bet the payoff is higher. Same applies to learning anything. Encouragement is underrated. I watch him while I exercise and drink my coffee. I say "beautiful" even when he messes it up (I got that from Richard Williams, Venus and Serena's dad). I celebrate his PRs. I take videos and I chart his daily progress so he can see the wins. Three months ago he could get 3 at his max, now his average is over 30. Seeing this progress over time has made it easier for him to show up and do the work every day. One focus at a time. We interleave other soccer skills (like shooting and dribbling), but tend to hyper focus on one area at a time so that we see fast progress. We use the same principle in our homeschooling. We are always working on a few things at once, but there is only one priority at a time. Start easy to build momentum. When you start your kids on a skill, you just want to be able to get *any* success at all. Don't be afraid to make it so easy it almost doesn't seem like the same skill. When we started I had him let the ball bounce between each kick. If I didn't do that it would have been so much harder for him to do keep uppy that he might have just quit. You want kids to get momentum in any skill, and you shouldn't be afraid to make the first steps so easy it almost doesn't seem like even the same skill. Focus on fundamentals. You might ask if keepy uppy is even important, since the real game is played mostly on the ground. The reason is found in the slight shuffling of his feet between each hit. In the real game, the quality of your control touch, shooting, and dribbling primarily comes down to these micro foot movements and how they position your body for balance. Keepy uppy trains the nervous system to a level of precision that is otherwise hard to attain. For education, the analogy is probably to something like reading. Reading is fundamental and makes you better at learning everything else, even if what you're reading about is unrelated to what you'll learn later. Students today would probably be better off just reading more vs. doing most of what passes today for "education." Beware of thresholds. The best people at keep uppy are not pro footballers. Every skill has a threshold where getting better at it confers little extra benefit. In education, I think the most obvious example is math. Everyone needs to master some fundmentals, but beyond it's rare to become more successful by getting better at math. Your effort is better spent on some complementary skill where you are weaker. Dress for success. Just kidding. I don't know why he's dressed like a ninja. That's all for now. Would love to hear how you guys think about building skills in the comments. 🙏

Chrisman

80,718 görüntüleme • 2 yıl önce