正在加载视频...

视频加载失败

加载此视频时出现问题。这可能是由于临时网络问题，或视频可能不可用。

trying to better visualize my coding sessions: each sphere = a project particles orbiting it are its tokens (1 particle ~ 10k) distance from center: last modified

Kat ⊷ the Poet Engineer

86,562 subscribers

19,618 次观看 • 4 天前 •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

0 条评论

暂无评论

原始帖子的评论将显示在这里

相关视频

Just asked the four Grok 4.20 (Beta) agents to build me an artificial life simulator. My mind is blown. What should I get him to build next? >>> grok prompt Build an artificial life simulator in one single HTML file. Hundreds of glowing neon particles in different cyberpunk colors. Each pair of colors has its own attraction/repulsion force that you control with sliders. You get emergent “creatures”, swarms, orbiting rings, exploding clusters, chasing packs, it feels alive. Features: - Click/drag to spawn particles of selected color - Real-time force matrix (6×6 sliders) - Randomize rules button (insane results every time) - Pause, speed, particle count, trails/glow toggle - Dark cyberpunk UI with scanlines + glitch effects optional

Just asked the four Grok 4.20 (Beta) agents to build me an artificial life simulator. My mind is blown. What should I get him to build next? >>> grok prompt Build an artificial life simulator in one single HTML file. Hundreds of glowing neon particles in different cyberpunk colors. Each pair of colors has its own attraction/repulsion force that you control with sliders. You get emergent “creatures”, swarms, orbiting rings, exploding clusters, chasing packs, it feels alive. Features: - Click/drag to spawn particles of selected color - Real-time force matrix (6×6 sliders) - Randomize rules button (insane results every time) - Pause, speed, particle count, trails/glow toggle - Dark cyberpunk UI with scanlines + glitch effects optional

tetsuo

360,532 次观看 • 6 个月前

Chainlink overhauls its Build program, steering value to LINK Chainlink just rewired how its Build program works. Rather than taking native tokens from the startups it supports, Chainlink Labs will now strike commercial deals paid in ethereum:0x514910771af9ca656af840dff83e8264ecf986ca, or in liquid assets converted to LINK, with proceeds funneled into the Chainlink Reserve. Stakers used to get free project tokens through Chainlink Rewards, about $20 million worth over time. That's ending now. The current season is the last, and claims close July 7. Chainlink frames it as "sustainable network economics," concentrating value in LINK rather than a basket of startup tokens.

Chainlink overhauls its Build program, steering value to LINK Chainlink just rewired how its Build program works. Rather than taking native tokens from the startups it supports, Chainlink Labs will now strike commercial deals paid in ethereum:0x514910771af9ca656af840dff83e8264ecf986ca, or in liquid assets converted to LINK, with proceeds funneled into the Chainlink Reserve. Stakers used to get free project tokens through Chainlink Rewards, about $20 million worth over time. That's ending now. The current season is the last, and claims close July 7. Chainlink frames it as "sustainable network economics," concentrating value in LINK rather than a basket of startup tokens.

BSCN

38,651 次观看 • 1 个月前

🚨 | Red Bull’s reverse wing is a separate project from Ferrari’s Macarena wing. The team submitted it to the FIA last summer, shortly after Ferrari, and developed it for months before its debut. 📰 AutoRacer IT

🚨 | Red Bull’s reverse wing is a separate project from Ferrari’s Macarena wing. The team submitted it to the FIA last summer, shortly after Ferrari, and developed it for months before its debut. 📰 AutoRacer IT

Race+

158,427 次观看 • 3 个月前

Ajith in an Interview to RP says, "Some who are trying to turn my interview against Vijay, its better for them to stay away from doing this. I have always meant the best for Vijay and wished him good things."

Ajith in an Interview to RP says, "Some who are trying to turn my interview against Vijay, its better for them to stay away from doing this. I have always meant the best for Vijay and wished him good things."

Mithun K Raman

16,653 次观看 • 9 个月前

Been exploring a new way to explore AI research papers to discover deeper insights. Agents are at the center of it. So far, I've built this little interactive artifact generator in my orchestrator to visualize things. This allows me to change views and insights (on-demand) from 100s of papers. Just scratching the surface here. More to share soon.

Been exploring a new way to explore AI research papers to discover deeper insights. Agents are at the center of it. So far, I've built this little interactive artifact generator in my orchestrator to visualize things. This allows me to change views and insights (on-demand) from 100s of papers. Just scratching the surface here. More to share soon.

elvis

141,678 次观看 • 5 个月前

Claude 3.7 Sonnet is king. I was having trouble adding mobile controls to my bolt.new project last night but with the latest update, it created these glass-like controls in one go. Try it here: I once spent a week manually coding similar controls 🥲

Claude 3.7 Sonnet is king. I was having trouble adding mobile controls to my bolt.new project last night but with the latest update, it created these glass-like controls in one go. Try it here: I once spent a week manually coding similar controls 🥲

Ian Curtis

49,633 次观看 • 1 年前

Went back to my swimming lessons today! (Last went in May) 🏊🏽‍♀️❤️ Trying out a new school and I think I like it better. Even the coach is so good.

Went back to my swimming lessons today! (Last went in May) 🏊🏽‍♀️❤️ Trying out a new school and I think I like it better. Even the coach is so good.

Noeleen

58,110 次观看 • 10 个月前

Trying out the new aglo… 🥺 Hiyaaa! I’m Diia! I’m a Fennec Vtuber who’s been on a 1 year project creating most of my assets from scratch, from my emotes to starting screens, down to my own Vtuber Model! If you’re interested in my journey consider following this fennec! 🍓✨

Trying out the new aglo… 🥺 Hiyaaa! I’m Diia! I’m a Fennec Vtuber who’s been on a 1 year project creating most of my assets from scratch, from my emotes to starting screens, down to my own Vtuber Model! If you’re interested in my journey consider following this fennec! 🍓✨

Diiaphy || DEBUT MAY 16 || 🍓✨

12,806 次观看 • 8 个月前

I revisited my #UE5 blackhole project recently. Updated the lighting. Before the light was only coming from the photon sphere around the event horizon. Now, in addition to that, I've added light coming from the gas itself, which gives it a much more realistic appearance. More screenshots in thread.

I revisited my #UE5 blackhole project recently. Updated the lighting. Before the light was only coming from the photon sphere around the event horizon. Now, in addition to that, I've added light coming from the gas itself, which gives it a much more realistic appearance. More screenshots in thread.

Overdraw Games

61,833 次观看 • 4 个月前

⚡️ 🇱🇹 Drone from Belarus enters Lithuanian airspace, spotted near Vilnius Local residents reported seeing the device in the sky around 5 a.m. According to Lithuanian authorities, the drone most likely came from Belarus. 💬 "We're trying to determine its current location and where it might have fallen," said Darius Buta, a spokesperson for the National Crisis Management Center. The UAV was last seen near Vilnius.

⚡️ 🇱🇹 Drone from Belarus enters Lithuanian airspace, spotted near Vilnius Local residents reported seeing the device in the sky around 5 a.m. According to Lithuanian authorities, the drone most likely came from Belarus. 💬 "We're trying to determine its current location and where it might have fallen," said Darius Buta, a spokesperson for the National Crisis Management Center. The UAV was last seen near Vilnius.

NEXTA

33,098 次观看 • 1 年前

Introducing Dexter 2.0 Open source. Built for financial research. Like Claude Code, but for stocks. What Dexter does: • plans tasks • runs on its own • validates its work • researches stocks It uses OSS tools like , with a fresh stack of typescript, react, ink, and bun. Fast to build. Easy to use. No coding required. My favorite project of the year.

Introducing Dexter 2.0 Open source. Built for financial research. Like Claude Code, but for stocks. What Dexter does: • plans tasks • runs on its own • validates its work • researches stocks It uses OSS tools like , with a fresh stack of typescript, react, ink, and bun. Fast to build. Easy to use. No coding required. My favorite project of the year.

Virat Singh

210,062 次观看 • 8 个月前

Here’s my shitty swing 😭 Just a reminder to all of you who are trying to make a change at full speed, chances are you couldn’t exaggerate the movement enough if you tried! I’m literally trying to get my hands above my head… Not even close. My swing goes from flat, to a little less flat, but not upright Also, I don’t put in enough reps to change it. You have a much better chance at changing your swing than I do I coach all day, you lot practice 🙂😂

Here’s my shitty swing 😭 Just a reminder to all of you who are trying to make a change at full speed, chances are you couldn’t exaggerate the movement enough if you tried! I’m literally trying to get my hands above my head… Not even close. My swing goes from flat, to a little less flat, but not upright Also, I don’t put in enough reps to change it. You have a much better chance at changing your swing than I do I coach all day, you lot practice 🙂😂

Ryan Mouque

85,288 次观看 • 1 个月前

I designed these worksheets to turn Agentic AI concepts into simple math problems you can do by hand. ✍️ Download PDF: Problems 1 to 5: 1. Count the tokens: split on spaces, one word per box 2. Subword splitting: when one word is three tokens 3. Punctuation counts: the marks are tokens too 4. Tokens per word: the ratio that turns words into a bill 5. Will it fit? A document against a context window Why am I making these worksheets? AI is making people (including me) think less. It's just too easy to ask AI a few questions about a new AI concept and start to believe I get it. Until recently, we could use a coding problem to practice a new AI concept. But now it has become too easy to ask AI to write the code, and we start to think we must know the concept, since we technically solved a coding problem. Thus, my approach is to recast AI concepts as simple math problems we must solve with pen and paper. Using our hands is one way to motivate ourselves to start thinking again. ✍️ ~ Prof. Tom Yeh

I designed these worksheets to turn Agentic AI concepts into simple math problems you can do by hand. ✍️ Download PDF: Problems 1 to 5: 1. Count the tokens: split on spaces, one word per box 2. Subword splitting: when one word is three tokens 3. Punctuation counts: the marks are tokens too 4. Tokens per word: the ratio that turns words into a bill 5. Will it fit? A document against a context window Why am I making these worksheets? AI is making people (including me) think less. It's just too easy to ask AI a few questions about a new AI concept and start to believe I get it. Until recently, we could use a coding problem to practice a new AI concept. But now it has become too easy to ask AI to write the code, and we start to think we must know the concept, since we technically solved a coding problem. Thus, my approach is to recast AI concepts as simple math problems we must solve with pen and paper. Using our hands is one way to motivate ourselves to start thinking again. ✍️ ~ Prof. Tom Yeh

Tom Yeh

18,291 次观看 • 11 天前

Finally getting to share one of my favorite projects. ICLR Oral! 🏆 It’s so strange how rigid video tokenization is. Think about it: why should a still landscape cost the same amount of tokens as a busy street? We built InfoTok. We went back to basics with Shannon’s information theory to make tokens "adaptive" in a principled way. Its 2.3x better compression and 11x faster inference demonstrates the magic of the old-school theory ✨ Check it out:

Finally getting to share one of my favorite projects. ICLR Oral! 🏆 It’s so strange how rigid video tokenization is. Think about it: why should a still landscape cost the same amount of tokens as a busy street? We built InfoTok. We went back to basics with Shannon’s information theory to make tokens "adaptive" in a principled way. Its 2.3x better compression and 11x faster inference demonstrates the magic of the old-school theory ✨ Check it out:

Haotian Ye

49,665 次观看 • 4 个月前

Motion graphics created by using Seedance 2.0 on Nsketch AI Prompt ⤵️ Ultra-sleek motion design sequence. Pure black background throughout. White-on-black minimal aesthetic. No cuts — one single continuous unbroken visual evolution. Locked-off static center frame, no camera movement. All motion occurs within the subject. Self-luminous white forms on absolute black, no grain, no shadows. 0–1s [ORIGIN]: A single white point materializes at center frame. It pulses once — three concentric rings ripple outward like a heartbeat and dissolve into black. 1–2s [LINE]: The point stretches into a razor-thin white horizontal line extending edge to edge with elastic ease-out motion. 2–4s [GRID → CIRCLE]: The line multiplies into a precise grid of nine parallel horizontal and vertical lines appearing in staggered succession. The grid warps — outer lines curve inward, the structure rotates, straight lines bend into arcs, converging into a perfect geometric circle outline. 4–6s [WIREFRAME → SOLID SPHERE]: The circle extrudes into a 3D wireframe sphere, every latitude and longitude line crisp, slowly rotating on its vertical axis. The wireframe fills in smoothly — dissolving into a solid white surface with a soft specular highlight in the upper-left quadrant. 6–8s [INTERFACE]: The sphere flattens and unfolds into a clean rectangular plane — a minimal phone interface with rounded corners. Abstract UI elements appear inside: status bar, a grid of rounded app icons, home indicator bar. White line art on black. 8–10s [CUBE → BLOOM]: The interface folds inward, collapsing into an isometric wireframe cube that tumbles once. The cube's six faces peel outward simultaneously, each becoming a smooth white elliptical petal — blooming open like a flower in zero gravity. 10–12s [VORTEX]: The flower dissolves into hundreds of tiny white particles orbiting the center in a controlled elliptical vortex, some clockwise, some counterclockwise, each following a precise mathematical path, flickering softly. 12–13s [WATCH]: The particles decelerate and self-organize — converging and snapping into position to assemble the silhouette of a wristwatch: a clean rectangular case outline with lugs, built entirely from particles. 13–14s [COLLAPSE]: The watch dissolves into fluid white ribbons that compress into a single dense white sphere. It pulses once — brighter than before — then collapses rapidly inward to a single brilliant point. 14–15s [LOGO]: The point holds for one silent beat. Then it expands into a clean minimal brand logo — white sans-serif letterforms on black. A single pulse of warm amber light blooms from center behind the logo, fades to black. White text remains. Style: luxury tech brand motion identity, Bauhaus-minimalist, mathematical precision. Mood: calm, inevitable, intelligent, premium.

Oogie

331,024 次观看 • 4 个月前

/1 Gemma 4 31B just crushed Qwen 3.6 27B in a local LLM gamedev contest inside atomic.chat (prompt is below) Device: MacBook Pro M5 Max, 64GB RAM Results: Qwen 3.6 27B: 32 tokens/sec · 18m 04s · 33,946 tokens Gemma 4 31B: 27 tokens/sec · 3m 51s · 6,209 tokens So what is more important: tokens per second, or the quality of the final answer? Qwen made a very long response and showed more creativity and visual style. But Gemma gave a shorter, clearer, and more logical answer in much less time. In this one-shot Pac-Man gamedev contest, Gemma 4 31B was the clear winner. Its game logic was stronger: click reactions were smoother, and it handled interactions with elements like walls, ghosts, and particle effects better. But this was only one test. Maybe Qwen 3.6 27B can show better results with better settings. Open the comments, try our prompt, and share your result below.

/1 Gemma 4 31B just crushed Qwen 3.6 27B in a local LLM gamedev contest inside atomic.chat (prompt is below) Device: MacBook Pro M5 Max, 64GB RAM Results: Qwen 3.6 27B: 32 tokens/sec · 18m 04s · 33,946 tokens Gemma 4 31B: 27 tokens/sec · 3m 51s · 6,209 tokens So what is more important: tokens per second, or the quality of the final answer? Qwen made a very long response and showed more creativity and visual style. But Gemma gave a shorter, clearer, and more logical answer in much less time. In this one-shot Pac-Man gamedev contest, Gemma 4 31B was the clear winner. Its game logic was stronger: click reactions were smoother, and it handled interactions with elements like walls, ghosts, and particle effects better. But this was only one test. Maybe Qwen 3.6 27B can show better results with better settings. Open the comments, try our prompt, and share your result below.

Chubby♨️

72,566 次观看 • 3 个月前

Truck drivers are trying the Tesla Semi. They love it. Center seat. You actually see everything. No leaning, no reaching. No engine shaking. No diesel fumes. Just a clean, quiet workspace. Cameras everywhere. Screens that actually make sense. That’s a better way to drive. Tesla, Eric Daugherty

Truck drivers are trying the Tesla Semi. They love it. Center seat. You actually see everything. No leaning, no reaching. No engine shaking. No diesel fumes. Just a clean, quiet workspace. Cameras everywhere. Screens that actually make sense. That’s a better way to drive. Tesla, Eric Daugherty

Mario Nawfal

77,426 次观看 • 4 个月前

🚨 BREAKING What if mass… isn’t a fixed property? Scientists just found evidence that mass can change. Not in theory. Inside real matter. They detected a rare state called an η′-mesic nucleus where a particle gets trapped inside a nucleus… and its mass shifts. Let that sink in. Same particle. Same physics. Different environment… different mass. That means: Mass isn’t just “what something is.” It’s how it interacts with the vacuum around it. And the vacuum? It’s not empty. It’s a hidden structure that gives particles their properties. Mass may come from the vacuum The vacuum changes in dense matter So mass itself can change This is huge. Because it flips the idea that mass is fundamental. Instead: mass might be emergent. In my view: Mass isn’t a property… it’s a response to time/field structure. Change the structure… you change the mass. So the real question is: Are particles “things”… or just stable patterns in a deeper field? Follow me physics isn’t discovering matter… it’s discovering what creates it.

🚨 BREAKING What if mass… isn’t a fixed property? Scientists just found evidence that mass can change. Not in theory. Inside real matter. They detected a rare state called an η′-mesic nucleus where a particle gets trapped inside a nucleus… and its mass shifts. Let that sink in. Same particle. Same physics. Different environment… different mass. That means: Mass isn’t just “what something is.” It’s how it interacts with the vacuum around it. And the vacuum? It’s not empty. It’s a hidden structure that gives particles their properties. Mass may come from the vacuum The vacuum changes in dense matter So mass itself can change This is huge. Because it flips the idea that mass is fundamental. Instead: mass might be emergent. In my view: Mass isn’t a property… it’s a response to time/field structure. Change the structure… you change the mass. So the real question is: Are particles “things”… or just stable patterns in a deeper field? Follow me physics isn’t discovering matter… it’s discovering what creates it.

TheNewPhysics

17,272 次观看 • 4 个月前

Claude Code Desktop now opens a new window for each session This makes it much easier to visualize multiple Claude Code agents running in parallel My current stack depends on the task: - Ghostty: when starting a project. Bash commands, git, env variables, provider connections. All manual through the terminal with a Claude panel running alongside. - Claude Code Desktop: once everything is configured. GitHub connected, CLAUDE.md, Skills, subagents and Hooks ready. Claude Code runs on its own, no more terminal setup, just panels running and outputs to review. - VSCode: when I need to review code by hand. I use it less and less, but there are moments where I have to confirm Claude got it right. I usually open the Claude extension inside VSCode, but it lacks most of the CLI features so it's limited Solid update. Worth trying once your workflows are already set up 👇

Claude Code Desktop now opens a new window for each session This makes it much easier to visualize multiple Claude Code agents running in parallel My current stack depends on the task: - Ghostty: when starting a project. Bash commands, git, env variables, provider connections. All manual through the terminal with a Claude panel running alongside. - Claude Code Desktop: once everything is configured. GitHub connected, CLAUDE.md, Skills, subagents and Hooks ready. Claude Code runs on its own, no more terminal setup, just panels running and outputs to review. - VSCode: when I need to review code by hand. I use it less and less, but there are moments where I have to confirm Claude got it right. I usually open the Claude extension inside VSCode, but it lacks most of the CLI features so it's limited Solid update. Worth trying once your workflows are already set up 👇

Daniel San

38,806 次观看 • 2 个月前

Congratulations 🍾🎉 to me It took me 5 sold years to complete this project, my mate are buying flashy cars I choose to do something better Finally a landlord in Abuja 😭 Thanks to CRYPTO , Massive giveaways to celebrate 🎉 this win 😭

Congratulations 🍾🎉 to me It took me 5 sold years to complete this project, my mate are buying flashy cars I choose to do something better Finally a landlord in Abuja 😭 Thanks to CRYPTO , Massive giveaways to celebrate 🎉 this win 😭

SOLDIER 🎖️

350,446 次观看 • 5 个月前