The Bugged Dev's banner

The Bugged Dev

@thebuggeddev • 6,221 subscribers

A Human Bugged by Bugs | Software Developer | Design Engineer | Building with AI | Learning Three.js 🥳

Shorts

Burned 200k tokens just for "animations" and still AI couldn't do it well. I tried one of the expert animation design engineer's, Emil Kowalski's Agent Skill on my website to improve animation using Grok 4.5 It analyzed the UI, found issues, came up with a good plan, and iterated multiple times. The end result actually made parts of the UX worse. Also, it never suggested the animation experience I had in mind. So I explained exactly how I wanted it to feel. It got much closer, but still missed details like applying the same animation during keyboard navigation between cards. That silly. The lesson isn't about the model or the Agent Skill. But it's about, "Even the strongest models can't infer your taste yet." They can't and won't suggest it. AI only executes. You have to define the experience with enough details for it to execute. Otherwise, you'll end up with something that's almost okay, but still not what you wanted

Burned 200k tokens just for "animations" and still AI couldn't do it well. I tried one of the expert animation design engineer's, Emil Kowalski's Agent Skill on my website to improve animation using Grok 4.5 It analyzed the UI, found issues, came up with a good plan, and iterated multiple times. The end result actually made parts of the UX worse. Also, it never suggested the animation experience I had in mind. So I explained exactly how I wanted it to feel. It got much closer, but still missed details like applying the same animation during keyboard navigation between cards. That silly. The lesson isn't about the model or the Agent Skill. But it's about, "Even the strongest models can't infer your taste yet." They can't and won't suggest it. AI only executes. You have to define the experience with enough details for it to execute. Otherwise, you'll end up with something that's almost okay, but still not what you wanted

23,256 Aufrufe

Sorry marcel, just vibe coded your cleanest dashboard ever in < 30 minutes 😅 Remix:

Sorry marcel, just vibe coded your cleanest dashboard ever in < 30 minutes 😅 Remix:

79,283 Aufrufe

OpenAI’s Sites feature is absolutely insane 🔥 I literally typed, “Build a content management website.” And some 10-15 minutes later, it had built it, configured it, managed git, and even deployed it without any errors or issues. P.S., This is mostly the UI with one round of design polishing, but adding the remaining functionality feels pretty straightforward! Live:

OpenAI’s Sites feature is absolutely insane 🔥 I literally typed, “Build a content management website.” And some 10-15 minutes later, it had built it, configured it, managed git, and even deployed it without any errors or issues. P.S., This is mostly the UI with one round of design polishing, but adding the remaining functionality feels pretty straightforward! Live:

45,749 Aufrufe

Wait, is Claude Opus 4.8 actually a Fable 5?? Opus generated this. A 3D botanical DNA strand. No, it's not a video or a 3D asset. It's pure math using Three.js, running at 60fps ❤️‍🔥 Everything including petals and flowers is generated using custom geometry. Also responsive. That too done from a image reference. Fable or not, Claude is generally good at 3D stuff.

Wait, is Claude Opus 4.8 actually a Fable 5?? Opus generated this. A 3D botanical DNA strand. No, it's not a video or a 3D asset. It's pure math using Three.js, running at 60fps ❤️‍🔥 Everything including petals and flowers is generated using custom geometry. Also responsive. That too done from a image reference. Fable or not, Claude is generally good at 3D stuff.

23,106 Aufrufe

Vibe coded an infinite 3D tunnel scroll using Three.js with object pooling and GSAP animations, all created with Google Gemini 3 Pro AI in under 30 minutes :D It’s a fully procedural, mathematically generated world that renders endlessly as you scroll up and down without eating your device memory (that's what at least AI said 😆) Live: Code: now returning to the job that funds my ability to overwork myself :)

Vibe coded an infinite 3D tunnel scroll using Three.js with object pooling and GSAP animations, all created with Google Gemini 3 Pro AI in under 30 minutes :D It’s a fully procedural, mathematically generated world that renders endlessly as you scroll up and down without eating your device memory (that's what at least AI said 😆) Live: Code: now returning to the job that funds my ability to overwork myself :)

95,858 Aufrufe

Ah yeah, just me trying to vibe code something new every day :D That 3D Rubik’s Cube you see there is built from scratch (custom math-based physics engine) using Three.js with GSAP animations and totally vibe coded in Google AI Studio using Gemini 3 Pro AI. Live: Code:

Ah yeah, just me trying to vibe code something new every day :D That 3D Rubik’s Cube you see there is built from scratch (custom math-based physics engine) using Three.js with GSAP animations and totally vibe coded in Google AI Studio using Gemini 3 Pro AI. Live: Code:

71,493 Aufrufe

Nah no, this is not a pre-made or baked Blender model. It is created entirely from scratch using extruded geometry and procedural textures for the mustard hatching pattern on the model... all done with Google Gemini 3 pro AI. In fact, every Three.js 3D model you see on this website has been vibe coded with AI 😎 Live: Code:

Nah no, this is not a pre-made or baked Blender model. It is created entirely from scratch using extruded geometry and procedural textures for the mustard hatching pattern on the model... all done with Google Gemini 3 pro AI. In fact, every Three.js 3D model you see on this website has been vibe coded with AI 😎 Live: Code:

54,391 Aufrufe

Claude feels kinda overhyped tbh. I tested the same website design prompt on both Gemini and Claude many times. And every time Claude’s output wasn’t even close to the original. Meanwhile, Gemini got almost identical results, and after a few tweaks, it hit ~99% accuracy. Makes me wonder… maybe models just perform better on ecosystems they’re trained closer to? 🤔 Either way, Gemini is seriously outperforming in frontend dev right now. Remix:

Claude feels kinda overhyped tbh. I tested the same website design prompt on both Gemini and Claude many times. And every time Claude’s output wasn’t even close to the original. Meanwhile, Gemini got almost identical results, and after a few tweaks, it hit ~99% accuracy. Makes me wonder… maybe models just perform better on ecosystems they’re trained closer to? 🤔 Either way, Gemini is seriously outperforming in frontend dev right now. Remix:

31,645 Aufrufe

Wrapping up my day by building a new biotech research lab website using Google Gemini 3 pro AI with immersive Three.js scenes shaders and more :D Live:

Wrapping up my day by building a new biotech research lab website using Google Gemini 3 pro AI with immersive Three.js scenes shaders and more :D Live:

46,971 Aufrufe

THIS IS ONE SHOTTED. ONE PROMPT. ONE RESULT. ☠️ Never ever imagined creating such slick 3D effects with Three.js would be this simple. I vibe coded with the latest Gemini Pro in Google AI Studio, and it’s honestly impressive how far these Gemini models have come in handling 3D. It feels like no other AI models come close to producing results like this :) Live: Code:

THIS IS ONE SHOTTED. ONE PROMPT. ONE RESULT. ☠️ Never ever imagined creating such slick 3D effects with Three.js would be this simple. I vibe coded with the latest Gemini Pro in Google AI Studio, and it’s honestly impressive how far these Gemini models have come in handling 3D. It feels like no other AI models come close to producing results like this :) Live: Code:

19,564 Aufrufe

Videos

Anya Rossi

sweetdream.ai

SweetDream.ai•Sponsored•Livecam

Watch Anya Live

Anya is streaming live right now! Join her private show and enjoy exclusive content.

Exclusive private shows

1.2k viewers online

Private Show

Join now for exclusive access

Free preview available • Premium content

Why do we still buy stadium tickets without ever seeing what the view from the seat actually looks like? ☠️ So I prototyped an idea. A 3D seat view experience for a football stadium built with Fable 5 + Three.js, where you can preview exactly what you'll see from your seat before buying. The entire 3D experience was working after the first prompt, and the whole prototype came together in just five prompts. This is the kind of 3D experience I'd love to see more of, tbh. Feels so good. Right now, buying a ticket gives you almost no sense of what the experience will actually feel like. We're still relying on static seat maps, charts, and boring UI layouts, when interactive 3D experiences could make that decision so much easier. If you think, 3D development has the potential to create genuinely useful experiences. This feels like the direction we should be heading. 3D is the future. I can already imagine this for cinemas, cricket stadiums, Wimbledon, concerts, and so many other venues. I'm planning to take this further. What should I build next? More stadium features, or should I try another venue? Let me know :D Code: Live:

Why do we still buy stadium tickets without ever seeing what the view from the seat actually looks like? ☠️ So I prototyped an idea. A 3D seat view experience for a football stadium built with Fable 5 + Three.js, where you can preview exactly what you'll see from your seat before buying. The entire 3D experience was working after the first prompt, and the whole prototype came together in just five prompts. This is the kind of 3D experience I'd love to see more of, tbh. Feels so good. Right now, buying a ticket gives you almost no sense of what the experience will actually feel like. We're still relying on static seat maps, charts, and boring UI layouts, when interactive 3D experiences could make that decision so much easier. If you think, 3D development has the potential to create genuinely useful experiences. This feels like the direction we should be heading. 3D is the future. I can already imagine this for cinemas, cricket stadiums, Wimbledon, concerts, and so many other venues. I'm planning to take this further. What should I build next? More stadium features, or should I try another venue? Let me know :D Code: Live:

1,599,696 Aufrufe • vor 2 Tagen

Eh, i'm really not impressed by GPT 5.6 sol lol. Asked it to recreate a 3 screen app design with image generation skill. Given how well it performs on benchmarks, I expected a lot more but the results were pretty ehhh. Not even close. Maybe I need to try harder but then... wasn't it supposed to do better?

Eh, i'm really not impressed by GPT 5.6 sol lol. Asked it to recreate a 3 screen app design with image generation skill. Given how well it performs on benchmarks, I expected a lot more but the results were pretty ehhh. Not even close. Maybe I need to try harder but then... wasn't it supposed to do better?

142,534 Aufrufe • vor 9 Tagen

This is one of the wildest things I have built with AI so far 😱 Asked Claude Fable 5 to create a fully interactive 3D crime investigation board using Three.js and it did it in ONE Prompt. It took one more prompt to give it a polished UI inspired by a reference design that I created with GPT Image. What surprised me most was not how it looked. It was how it worked. Fable built a complete Verlet rope physics system with 46 particles in every thread. It handled gravity, damping, distance constraints and all other features of file upload, text and everything. The threads/ropes even lift naturally with pinned evidence instead of clipping through it. I never asked it to implement any of that. I only described the experience I wanted. Maybe that's how we should work with AI. Instead of writing long long specs, we should explain what we want as an experience and let AI do its thing. Anyway, every day Fable 5 raises the bar for what I think AI can build. More to this Live: Code:

This is one of the wildest things I have built with AI so far 😱 Asked Claude Fable 5 to create a fully interactive 3D crime investigation board using Three.js and it did it in ONE Prompt. It took one more prompt to give it a polished UI inspired by a reference design that I created with GPT Image. What surprised me most was not how it looked. It was how it worked. Fable built a complete Verlet rope physics system with 46 particles in every thread. It handled gravity, damping, distance constraints and all other features of file upload, text and everything. The threads/ropes even lift naturally with pinned evidence instead of clipping through it. I never asked it to implement any of that. I only described the experience I wanted. Maybe that's how we should work with AI. Instead of writing long long specs, we should explain what we want as an experience and let AI do its thing. Anyway, every day Fable 5 raises the bar for what I think AI can build. More to this Live: Code:

96,025 Aufrufe • vor 8 Tagen

Everyone is using Fable 5 to build funny little demos so I decided to build something better. Vibe coded a 3D keyboard ecom platform in Three.js and GSAP. Every key. Every switch. Every color. Customizable in real time in 3D. It's really amusing that whole keyboard is built at runtime using Procedural geometry. To achieve that kind of realistic lighting on keyboard, it used a studio environment (gradient sky + softbox panels) rendered into a PMREM cubemap for realistic reflections. Opus helped a bit more here (as Fable was failing in loop in Claude chat) There is also custom GLSL shaders for RGB underflow/skirt and as well sRGB output encoding + explicit convertSRGBToLinear() on every material color so caps render at their exact swatch value. I don't know any of these techniques and concepts but I have had been reading how both Fable and Opus were trying to approach this by digging into their "thinking" phase. Go try it, there is a cool Claude theme keyboard you would want to purchase too :D Live: Code:

Everyone is using Fable 5 to build funny little demos so I decided to build something better. Vibe coded a 3D keyboard ecom platform in Three.js and GSAP. Every key. Every switch. Every color. Customizable in real time in 3D. It's really amusing that whole keyboard is built at runtime using Procedural geometry. To achieve that kind of realistic lighting on keyboard, it used a studio environment (gradient sky + softbox panels) rendered into a PMREM cubemap for realistic reflections. Opus helped a bit more here (as Fable was failing in loop in Claude chat) There is also custom GLSL shaders for RGB underflow/skirt and as well sRGB output encoding + explicit convertSRGBToLinear() on every material color so caps render at their exact swatch value. I don't know any of these techniques and concepts but I have had been reading how both Fable and Opus were trying to approach this by digging into their "thinking" phase. Go try it, there is a cool Claude theme keyboard you would want to purchase too :D Live: Code:

95,204 Aufrufe • vor 14 Tagen

Finally Fable 5 was able to code and it cooked this in Three.js ❤️‍🔥 It's still crazy how it created all of these models purely using Math. All of these, with animations, effects, labels and such in just two prompts. More updates coming soon :)

Finally Fable 5 was able to code and it cooked this in Three.js ❤️‍🔥 It's still crazy how it created all of these models purely using Math. All of these, with animations, effects, labels and such in just two prompts. More updates coming soon :)

81,933 Aufrufe • vor 17 Tagen

Couldn't get enough of Claude Fable 5, so I challenged it to build a fully 3D mechanical engine platform and it just did!! 😱 And no, these aren't pre made Blender models. Nah. Fable generated everything from scratch using custom geometry, built entirely using Three.js Zero assets. One prompt. Crazzzzy results

Couldn't get enough of Claude Fable 5, so I challenged it to build a fully 3D mechanical engine platform and it just did!! 😱 And no, these aren't pre made Blender models. Nah. Fable generated everything from scratch using custom geometry, built entirely using Three.js Zero assets. One prompt. Crazzzzy results

82,068 Aufrufe • vor 1 Monat

ChatGPT Image 2.0 created the design. Codex GPT 5.5 turned that design into a full Expo React Native app From asset generation to music logic, everything handled by ChatGPT using just a referenced design image 🔥 Is this what the future of development looks like?

ChatGPT Image 2.0 created the design. Codex GPT 5.5 turned that design into a full Expo React Native app From asset generation to music logic, everything handled by ChatGPT using just a referenced design image 🔥 Is this what the future of development looks like?

111,376 Aufrufe • vor 2 Monaten

For anyone who believes AI cannot build multi page complex applications, here is my proof. A fully functional 20 page dashboard created with Gemini 3.1 Pro inside Google AI Studio 🔥 It did not take a day or even half a day. It took me less than an hour to complete it in 17 total iterations. The UI design is inspired by Salung Prastyo and the rest pages were generated entirely by Gemini following same design system. More features coming soon 🔜

For anyone who believes AI cannot build multi page complex applications, here is my proof. A fully functional 20 page dashboard created with Gemini 3.1 Pro inside Google AI Studio 🔥 It did not take a day or even half a day. It took me less than an hour to complete it in 17 total iterations. The UI design is inspired by Salung Prastyo and the rest pages were generated entirely by Gemini following same design system. More features coming soon 🔜

167,235 Aufrufe • vor 4 Monaten

Holy moly, I finally vibe coded a website I’ve always dreamt of building 🥹 Nah, it’s not a Blender model. Not even for 3D podium. It’s pure geometry built using Three.js with color maps (albedo), spring-based physics, a hybrid DOM, real-time planar reflections, and more. Completely vibe coded all using Gemini 3 Pro in Google AI Studio!! Live:

Holy moly, I finally vibe coded a website I’ve always dreamt of building 🥹 Nah, it’s not a Blender model. Not even for 3D podium. It’s pure geometry built using Three.js with color maps (albedo), spring-based physics, a hybrid DOM, real-time planar reflections, and more. Completely vibe coded all using Gemini 3 Pro in Google AI Studio!! Live:

165,507 Aufrufe • vor 5 Monaten

Gemini built a learning management platform like Udemy with 10-12 pages all inside Google AI Studio 🔥 It took just a couple of hours. That's how accessible building complex websites is becoming. If you have an idea, you may not need other services now. With AI, You can build it yourself, your way, at your own scale :) Remix:

Gemini built a learning management platform like Udemy with 10-12 pages all inside Google AI Studio 🔥 It took just a couple of hours. That's how accessible building complex websites is becoming. If you have an idea, you may not need other services now. With AI, You can build it yourself, your way, at your own scale :) Remix:

55,438 Aufrufe • vor 2 Monaten

Can't believe it. Can't believe AI. I just vibe coded a diary with realistic page flip GSAP animations entirely using Gemini 3.1 Pro in Google AI Studio. And yeah… it’s fully responsive too. At this point what even is “development” anymore?? Live: Code:

Can't believe it. Can't believe AI. I just vibe coded a diary with realistic page flip GSAP animations entirely using Gemini 3.1 Pro in Google AI Studio. And yeah… it’s fully responsive too. At this point what even is “development” anymore?? Live: Code:

54,391 Aufrufe • vor 2 Monaten

Okay this one’s been living in my head for a while… finally built it ❤️‍🔥 A fully vibe coded interactive planet explorer. No blender models, just pure geometry, pure math, using Three.js with frustum, culling Logic, raycasting, spring physics, and more. Built using Gemini 3.1 Pro in Google AI Studio as always :D Live:

Okay this one’s been living in my head for a while… finally built it ❤️‍🔥 A fully vibe coded interactive planet explorer. No blender models, just pure geometry, pure math, using Three.js with frustum, culling Logic, raycasting, spring physics, and more. Built using Gemini 3.1 Pro in Google AI Studio as always :D Live:

44,138 Aufrufe • vor 3 Monaten

One Prompt. One Iteration. That’s all it took to vibe code this super sleek 3D orbital gallery using Three.js with Gemini 3.1 Pro in Google AI Studio. Period. Live: Code:

One Prompt. One Iteration. That’s all it took to vibe code this super sleek 3D orbital gallery using Three.js with Gemini 3.1 Pro in Google AI Studio. Period. Live: Code:

46,186 Aufrufe • vor 3 Monaten

$Inspired by Omma's project, vibe coded a generative voxel planet website using Three.js in Google AI Studio with Gemini 3.1 pro. Freakinn craziest thing is, this is PURE MATH. No 3D model and each planet renders 25,000 voxels cubes in fraction of second. Trying to build something like this manually would be an absolute nightmare for me 🫠 Live: Code:$

Inspired by Omma's project, vibe coded a generative voxel planet website using Three.js in Google AI Studio with Gemini 3.1 pro. Freakinn craziest thing is, this is PURE MATH. No 3D model and each planet renders 25,000 voxels cubes in fraction of second. Trying to build something like this manually would be an absolute nightmare for me 🫠 Live: Code:

46,944 Aufrufe • vor 3 Monaten

Jeeezz, I just vibe coded a fully interactive 3D solar system explorer using Three.js and GSAP in Google AI Studio with Gemini 3.1 Pro. Still can't believe it. It uses advanced threejs strategies like Hybird DOM, WebGL Masking, cinematic camera, Fresnel effect and such. Building stuff like this genuinely feels like joy. Smooth, immersive, and ridiculously fun to create using AI ♥️ It's inspired from Jamie Coulter original CSS solar explorer codepen :) Live: Code:

Jeeezz, I just vibe coded a fully interactive 3D solar system explorer using Three.js and GSAP in Google AI Studio with Gemini 3.1 Pro. Still can't believe it. It uses advanced threejs strategies like Hybird DOM, WebGL Masking, cinematic camera, Fresnel effect and such. Building stuff like this genuinely feels like joy. Smooth, immersive, and ridiculously fun to create using AI ♥️ It's inspired from Jamie Coulter original CSS solar explorer codepen :) Live: Code:

45,891 Aufrufe • vor 3 Monaten

I still can't believe I (AI) made this 😱 A flying physics based wheel carousel built from pure Three.js geometry. No models. No assets. Pure MATH!! Claude Opus 4.8 nailed it in almost one shot. Crazy what AI can do these days. Live: Code:

I still can't believe I (AI) made this 😱 A flying physics based wheel carousel built from pure Three.js geometry. No models. No assets. Pure MATH!! Claude Opus 4.8 nailed it in almost one shot. Crazy what AI can do these days. Live: Code:

10,479 Aufrufe • vor 23 Tagen

This is PURE MATH!! Built fully with Three.js and completely vibe coded using Gemini 3.1 Pro in Google AI Studio. This is what AI can do right now. I have been vibe coding a lot of websites lately and every day these models go beyond what I thought was possible 🤯 Still in awe. Expect more voxel 3D drops coming to my feed :) Remix:

This is PURE MATH!! Built fully with Three.js and completely vibe coded using Gemini 3.1 Pro in Google AI Studio. This is what AI can do right now. I have been vibe coding a lot of websites lately and every day these models go beyond what I thought was possible 🤯 Still in awe. Expect more voxel 3D drops coming to my feed :) Remix:

29,847 Aufrufe • vor 3 Monaten

AI can't build complex websites, they said. So I vibe coded a hyper realistic 3D tunnel gallery using Three.js with Gemini 3.1 pro in Google AI Studio in < 1 hour. It runs on Catmull Rom splines with a treadmill style sliding geometry, custom fragment shaders with depth based fog, and tight optimizations like texture disposal, disabled frustum checks, and capped DPR to keep it smooth even on long infinite scrolls. All of this is seriously hard to learn and build manually, but AI pulls it together so well! Live: Code:

AI can't build complex websites, they said. So I vibe coded a hyper realistic 3D tunnel gallery using Three.js with Gemini 3.1 pro in Google AI Studio in < 1 hour. It runs on Catmull Rom splines with a treadmill style sliding geometry, custom fragment shaders with depth based fog, and tight optimizations like texture disposal, disabled frustum checks, and capped DPR to keep it smooth even on long infinite scrolls. All of this is seriously hard to learn and build manually, but AI pulls it together so well! Live: Code:

18,762 Aufrufe • vor 3 Monaten

If you think a website can’t be vibe coded from just a design reference image, you’re seriously underestimating what’s possible 🥶 I’ve been building a lot of websites this way lately, and this is one of them. Vibe coded a MindInventory’s design using Gemini inside Google AI Studio in under 1 hour 🔥

If you think a website can’t be vibe coded from just a design reference image, you’re seriously underestimating what’s possible 🥶 I’ve been building a lot of websites this way lately, and this is one of them. Vibe coded a MindInventory’s design using Gemini inside Google AI Studio in under 1 hour 🔥

12,230 Aufrufe • vor 2 Monaten

Vibe coded something new today 😉 It’s a 3D Solana AI protocol website built using Gemini 3 Pro in Google AI Studio. All the Three.js models you see on the site have been built from scratch by AI using different geometries, PBR, Linear Interpolation, Lightformers (IBL), Shadows and more :D Live: Code:

Vibe coded something new today 😉 It’s a 3D Solana AI protocol website built using Gemini 3 Pro in Google AI Studio. All the Three.js models you see on the site have been built from scratch by AI using different geometries, PBR, Linear Interpolation, Lightformers (IBL), Shadows and more :D Live: Code:

21,128 Aufrufe • vor 6 Monaten