Loading video...

Video Failed to Load

There was a problem loading this video. This could be due to a temporary network issue or the video might be unavailable.

At this point, Three.js is probably the best graphics stack for AI-generated games. Not because it magically produces better visuals, but because models are extremely good at reasoning directly over JavaScript, scene graphs, shaders, procedural geometry, physics integrations, and fast browser iteration. The examples here make the case pretty... show more

Token Gremlin

5,831 subscribers

12,026 views • 24 days ago •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

$Fable 5 built this real-time ocean scene directly in the browser using Three.js, WebGPU and TSL, with no engine and no baked assets. The interesting part is not just that the water looks good. The whole system is being simulated: wave spectrum, GPU FFT, foam generation and refracted-ray caustics, all running live. This is exactly where frontier models start becoming genuinely useful for technical graphics work. Not by replacing years of rendering research, but by compressing the path from complex math to a working interactive result. A small, focused scene, but technically very serious. Cc: Adem Vessell$

Fable 5 built this real-time ocean scene directly in the browser using Three.js, WebGPU and TSL, with no engine and no baked assets. The interesting part is not just that the water looks good. The whole system is being simulated: wave spectrum, GPU FFT, foam generation and refracted-ray caustics, all running live. This is exactly where frontier models start becoming genuinely useful for technical graphics work. Not by replacing years of rendering research, but by compressing the path from complex math to a working interactive result. A small, focused scene, but technically very serious. Cc: Adem Vessell

Token Gremlin

21,731 views • 1 month ago

Astra built this three.js game with gorgeous graphics, then recreated every scene in Blender and made the trailer too. It feels like a massive upgrade from Sol for 3D graphics and front-end design. The UI feels way more polished and less like AI slop. The rocks, trees, monuments, and characters are all procedural code, keeping the game under 2 MB before the image and music assets. The level of autonomy is unbelievable. It needed very little steering, though I still had to keep pushing it to work harder on the details. The UI icons, buttons, and ornaments were made as transparent PNGs with GPT Image 2.

Astra built this three.js game with gorgeous graphics, then recreated every scene in Blender and made the trailer too. It feels like a massive upgrade from Sol for 3D graphics and front-end design. The UI feels way more polished and less like AI slop. The rocks, trees, monuments, and characters are all procedural code, keeping the game under 2 MB before the image and music assets. The level of autonomy is unbelievable. It needed very little steering, though I still had to keep pushing it to work harder on the details. The UI icons, buttons, and ornaments were made as transparent PNGs with GPT Image 2.

Meng To

129,821 views • 4 days ago

I'm playing around with generative AI tools and stitching them together into visual stories. Here I took the first few sentences of Pride and Prejudice and made it into a video. The gen stack used for this one: - Anthropic Claude took the first chapter, generated the scenes and the individual prompts to to the image generator. - Ideogram took the prompts and generate the images - Luma took the images and animated them - for narration - VEED | AI Video Creation to stitch it together (Many of these choices are just what I happened to use for this one while exploring a bunch of things). Anyway honestly it was pretty messy and there is a ton of copy pasting between all of the tools, and even this little video with 3 scenes took me about an hour. There is a huge storytelling opportunity here for whoever can make this convenient. Who is building the first 100% AI-native movie maker?

I'm playing around with generative AI tools and stitching them together into visual stories. Here I took the first few sentences of Pride and Prejudice and made it into a video. The gen stack used for this one: - Anthropic Claude took the first chapter, generated the scenes and the individual prompts to to the image generator. - Ideogram took the prompts and generate the images - Luma took the images and animated them - for narration - VEED | AI Video Creation to stitch it together (Many of these choices are just what I happened to use for this one while exploring a bunch of things). Anyway honestly it was pretty messy and there is a ton of copy pasting between all of the tools, and even this little video with 3 scenes took me about an hour. There is a huge storytelling opportunity here for whoever can make this convenient. Who is building the first 100% AI-native movie maker?

Andrej Karpathy

609,834 views • 2 years ago

It cost me $15 on the Epic Games Store and a few million AI tokens to build this web-native grass simulation. Demo in the comments. Epic Games owns Quixel, a collection of high-quality Megascans. You can use them in Three.js, which is exactly what we did in this demo. The butterflies, flowers, and surfaces all come from Megascans. The long grass and wind simulation were converted from Blender Geometry Nodes simulations, which is why it feels so good. Then add an HDRI, set up the lighting, and boom—you have highly realistic grass rendered fully on the web. While it’s something built with AI, a non-designer without a certain set of skills would never be able to do this. You still need the knowledge, terminology, and understanding required to glue all these technologies together.

It cost me $15 on the Epic Games Store and a few million AI tokens to build this web-native grass simulation. Demo in the comments. Epic Games owns Quixel, a collection of high-quality Megascans. You can use them in Three.js, which is exactly what we did in this demo. The butterflies, flowers, and surfaces all come from Megascans. The long grass and wind simulation were converted from Blender Geometry Nodes simulations, which is why it feels so good. Then add an HDRI, set up the lighting, and boom—you have highly realistic grass rendered fully on the web. While it’s something built with AI, a non-designer without a certain set of skills would never be able to do this. You still need the knowledge, terminology, and understanding required to glue all these technologies together.

Alex Barashkov

16,448 views • 1 month ago

I asked Astra to create a Catan game in three.js and it came out incredibly well. I played it, and the animations make it so much fun. The instructions and info cards make the game easy to learn. It even has AI players and trading. Astra used another game I had started building as a reference. The more games I build, the more skills and assets I have for the next one. It can remix the UI, 3D models, pages, and sounds instead of rebuilding everything. Now I want to see how many other board games I can make from this foundation.

I asked Astra to create a Catan game in three.js and it came out incredibly well. I played it, and the animations make it so much fun. The instructions and info cards make the game easy to learn. It even has AI players and trading. Astra used another game I had started building as a reference. The more games I build, the more skills and assets I have for the next one. It can remix the UI, 3D models, pages, and sounds instead of rebuilding everything. Now I want to see how many other board games I can make from this foundation.

Meng To

485,161 views • 1 day ago

Can we call this AI directing? 😂 Seedance 2.0 with the Omni Reference feature honestly feels like a total game changer. Summer Trip is one of my film which I made used Omni Reference, not only start frame anymore and for the first time, it genuinely feels like I’m directing the AI instead of just prompting it. The reasoning in Seedance 2.0 is seriously impressive. It understands intent in a way that feels collaborative. Of course, in production it still needs proper touch-ups in post, grading, timing, polishing details, but the foundation it gives is strong. Like it or not, we’re already here. And to be honest, my workflow has to evolve too. I can’t approach this the same way I did before. It’s less about generating random shots and more about guiding, refining, and thinking like a director from the start. Personally, I see this as a shift, not replacing filmmaking, but reshaping the role of the filmmaker. Now the question isn’t “Can AI make it?” but “How well can you direct it?”

Can we call this AI directing? 😂 Seedance 2.0 with the Omni Reference feature honestly feels like a total game changer. Summer Trip is one of my film which I made used Omni Reference, not only start frame anymore and for the first time, it genuinely feels like I’m directing the AI instead of just prompting it. The reasoning in Seedance 2.0 is seriously impressive. It understands intent in a way that feels collaborative. Of course, in production it still needs proper touch-ups in post, grading, timing, polishing details, but the foundation it gives is strong. Like it or not, we’re already here. And to be honest, my workflow has to evolve too. I can’t approach this the same way I did before. It’s less about generating random shots and more about guiding, refining, and thinking like a director from the start. Personally, I see this as a shift, not replacing filmmaking, but reshaping the role of the filmmaker. Now the question isn’t “Can AI make it?” but “How well can you direct it?”

Dinda Prasetyo

29,385 views • 6 months ago

Fable 5.1 is an absolute beast for game dev. I took my Opus 5 space game and told Fable to "just make it look better". It overhauled everything with a TON of detail, AND made it 2x faster 🤯 Code: Demo: Every pixel you see is made by AI. Fable modeled everything in Blender, generated textures with code, and in a few cases ran image models via CLI. In mere months we'll be vibe-coding AAA graphics at this rate. Just insane. Of course, Fable recorded and edited this comparison video too (with a little help from Opus at the end, because of the Fable outage)

Fable 5.1 is an absolute beast for game dev. I took my Opus 5 space game and told Fable to "just make it look better". It overhauled everything with a TON of detail, AND made it 2x faster 🤯 Code: Demo: Every pixel you see is made by AI. Fable modeled everything in Blender, generated textures with code, and in a few cases ran image models via CLI. In mere months we'll be vibe-coding AAA graphics at this rate. Just insane. Of course, Fable recorded and edited this comparison video too (with a little help from Opus at the end, because of the Fable outage)

Anshu

48,983 views • 5 days ago

Today, I had the privilege of speaking to the HUMAIN team at our CEO Townhall. I often say this, and I mean it deeply: I am living my dream... working in a country where energy goes beyond oil. It’s in the people, the ambition, and the belief in building a better future. Saudi Arabia is unlike anywhere else. The hospitality is second to none. The vision is bold. And the commitment to shaping the future is real. What we are building at HUMAIN is foundational. We are not only participating in the AI era, but redefining it. Shifting the narrative from experimentation to real value creation. But more importantly, we are not afraid to challenge what an organization should look like in an AI-native world. In fact, we are pioneering it. We are actively reshaping how we work: - Moving toward a future where AI agents do the work - Empowering our people to focus on high-value thinking, creativity, and decision-making - Building strong foundations in security, governance, and guardrails - Continuously enhancing our models, systems, and operating frameworks This is not theory. This is happening now. What I saw today from our teams gives me absolute confidence: - Products that are not just innovative, but game-changing - Teams building at a speed and quality that challenge global norms - A culture focused on execution, ownership, and impact At #LEAP2026 this year, we will go beyond vision. We will show: Real demos. Real products. Some of them are a first of their kind in the world. And they are built right here in Saudi Arabia. I could not be more proud of this team. What they have accomplished in such a short time is remarkable. This is just the beginning. The future is not something we wait for. It’s something we build. #HUMAIN #LEAP #TheEndOfLimits #AI

Today, I had the privilege of speaking to the HUMAIN team at our CEO Townhall. I often say this, and I mean it deeply: I am living my dream... working in a country where energy goes beyond oil. It’s in the people, the ambition, and the belief in building a better future. Saudi Arabia is unlike anywhere else. The hospitality is second to none. The vision is bold. And the commitment to shaping the future is real. What we are building at HUMAIN is foundational. We are not only participating in the AI era, but redefining it. Shifting the narrative from experimentation to real value creation. But more importantly, we are not afraid to challenge what an organization should look like in an AI-native world. In fact, we are pioneering it. We are actively reshaping how we work: - Moving toward a future where AI agents do the work - Empowering our people to focus on high-value thinking, creativity, and decision-making - Building strong foundations in security, governance, and guardrails - Continuously enhancing our models, systems, and operating frameworks This is not theory. This is happening now. What I saw today from our teams gives me absolute confidence: - Products that are not just innovative, but game-changing - Teams building at a speed and quality that challenge global norms - A culture focused on execution, ownership, and impact At #LEAP2026 this year, we will go beyond vision. We will show: Real demos. Real products. Some of them are a first of their kind in the world. And they are built right here in Saudi Arabia. I could not be more proud of this team. What they have accomplished in such a short time is remarkable. This is just the beginning. The future is not something we wait for. It’s something we build. #HUMAIN #LEAP #TheEndOfLimits #AI

Tareq Amin

17,856 views • 4 months ago

"I like to move extremely fast, but in order to move fast, you need to have brakes." Up to 200,000 AI agents wake up at Kavak every day. CPO & AI Officer Alejandro Maza says eval quality is what sets the ceiling on speed at that scale: "So to give you a sense of the scale, 96% of all interactions are handled by agents. No humans there. 95% of all transactions are completely handled by agents." "AI is super powerful, and I've seen many companies get this wrong because they try to go slow because they don't have the right brakes. So I thought about it the other way around: how fast can we go? Well, it depends on the quality of our evals." "A good rule of thumb here is we spend about the same amount of engineer time, tokens, and money on building the evals as building the agents." Alejandro Maza

"I like to move extremely fast, but in order to move fast, you need to have brakes." Up to 200,000 AI agents wake up at Kavak every day. CPO & AI Officer Alejandro Maza says eval quality is what sets the ceiling on speed at that scale: "So to give you a sense of the scale, 96% of all interactions are handled by agents. No humans there. 95% of all transactions are completely handled by agents." "AI is super powerful, and I've seen many companies get this wrong because they try to go slow because they don't have the right brakes. So I thought about it the other way around: how fast can we go? Well, it depends on the quality of our evals." "A good rule of thumb here is we spend about the same amount of engineer time, tokens, and money on building the evals as building the agents." Alejandro Maza

a16z

41,467 views • 29 days ago

I tested Rust vs Javascript for games 🎮 I had this question recently and decided to port my Three.js game to Rust Desktop I'm still not able to reproduce the same look on both games but I believe that codewise they are very similar With both games running, the Rust version averages 37FPS while the Three.js version averages around 20FPS If I'm running one game at a time the FPS goes up considerably (60 FPS on both with power on) So the Rust version for Desktop has 85% increase in performance on my Macbook Pro M4 This comparision is basically comparing Browser vs Native, which is not completely fair So that's why I want to port the Rust version to WASM on the browser and benchmark again soon

I tested Rust vs Javascript for games 🎮 I had this question recently and decided to port my Three.js game to Rust Desktop I'm still not able to reproduce the same look on both games but I believe that codewise they are very similar With both games running, the Rust version averages 37FPS while the Three.js version averages around 20FPS If I'm running one game at a time the FPS goes up considerably (60 FPS on both with power on) So the Rust version for Desktop has 85% increase in performance on my Macbook Pro M4 This comparision is basically comparing Browser vs Native, which is not completely fair So that's why I want to port the Rust version to WASM on the browser and benchmark again soon

André → andreelias.dev

93,523 views • 3 months ago

Boom! I have just been gifted 100s hours of war plane chatter across decides. It is quite intense and almost none of it has been digitalized and made public. I will digitize it and make it open source. I will be training YOUR AI in this high value data. It is not just valuable to the AI model for history but also thousands of hours of grace under pressure. These men had extremely high honor and dignity. You will not find this caliber of human interaction on Reddit interactions slurped up by OpenAI models. But you will find it in the AI that I build. — This radio chatter clip is from a Lancaster Bomber which is flying a mission over Germany during World War II. The bomber is attacked by a German fighter later on in the clip. The crew seems to panic as the fighter engages them, and the captain at one point shouts "Okay, don't shout all at once!" Eventually one of the gunners manages to bring down the German fighter.

Boom! I have just been gifted 100s hours of war plane chatter across decides. It is quite intense and almost none of it has been digitalized and made public. I will digitize it and make it open source. I will be training YOUR AI in this high value data. It is not just valuable to the AI model for history but also thousands of hours of grace under pressure. These men had extremely high honor and dignity. You will not find this caliber of human interaction on Reddit interactions slurped up by OpenAI models. But you will find it in the AI that I build. — This radio chatter clip is from a Lancaster Bomber which is flying a mission over Germany during World War II. The bomber is attacked by a German fighter later on in the clip. The crew seems to panic as the fighter engages them, and the captain at one point shouts "Okay, don't shout all at once!" Eventually one of the gunners manages to bring down the German fighter.

Brian Roemmele

24,544 views • 8 months ago

$This is running in a browser tab. No engine. No install. No download. Six spells, real-time particle simulation, refracting ice geometry, and a live VFX editor on the right side where every parameter is tunable while the scene runs. A studio VFX artist builds this in weeks. It is genuinely the hardest discipline in 3D - nothing about particle work is intuitive. The thing that changed is not that AI got better at art. It is that Astra writes geometry and shaders as CODE instead of as files. A game made of .fbx files is a game an AI can only rearrange. A game made of code is a game an AI can edit. That is why the good demos this week are all three.js and Blender-scripted, and why they run at 60fps in a tab. Dude built a full game this way and it came out under 2 MB before audio. Every rock, tree and character - procedural code, not assets. The cost consequence is the part almost nobody understood: → let the model CLICK in Blender: burns your weekly cap in two minutes → let the model WRITE Blender Python: ~100x cheaper One builder shipped a complete 3D roguelike level - day/night cycle, weather, modeling, texturing, enemies - for 3% of his quota. That is the entire reason why. Scripts, not clicks. It's the single rule that decides your bill. Full breakdown + the 7-day guide in the article below ↓$

This is running in a browser tab. No engine. No install. No download. Six spells, real-time particle simulation, refracting ice geometry, and a live VFX editor on the right side where every parameter is tunable while the scene runs. A studio VFX artist builds this in weeks. It is genuinely the hardest discipline in 3D - nothing about particle work is intuitive. The thing that changed is not that AI got better at art. It is that Astra writes geometry and shaders as CODE instead of as files. A game made of .fbx files is a game an AI can only rearrange. A game made of code is a game an AI can edit. That is why the good demos this week are all three.js and Blender-scripted, and why they run at 60fps in a tab. Dude built a full game this way and it came out under 2 MB before audio. Every rock, tree and character - procedural code, not assets. The cost consequence is the part almost nobody understood: → let the model CLICK in Blender: burns your weekly cap in two minutes → let the model WRITE Blender Python: ~100x cheaper One builder shipped a complete 3D roguelike level - day/night cycle, weather, modeling, texturing, enemies - for 3% of his quota. That is the entire reason why. Scripts, not clicks. It's the single rule that decides your bill. Full breakdown + the 7-day guide in the article below ↓

Aleiah

37,652 views • 2 days ago

I just watched AI build a game from a sentence. No Unity. No Unreal. No code. No files. I simply described the game I wanted to make, and Gear Zero planned it, coded it, created the visuals, and turned it into a playable game in my browser. But this is the part that got me 👇 You don’t just generate it and leave. You keep talking to Gear. Change things. Add features. Polish the game. Keep building. It feels less like using an AI game generator… …and more like having a game dev team you can talk to 24/7. The future of making games might actually look like this. Try it →

I just watched AI build a game from a sentence. No Unity. No Unreal. No code. No files. I simply described the game I wanted to make, and Gear Zero planned it, coded it, created the visuals, and turned it into a playable game in my browser. But this is the part that got me 👇 You don’t just generate it and leave. You keep talking to Gear. Change things. Add features. Polish the game. Keep building. It feels less like using an AI game generator… …and more like having a game dev team you can talk to 24/7. The future of making games might actually look like this. Try it →

RAVI KUMAR SAHU

30,938 views • 8 days ago

Sound ON/Founder Voice-Over: Inside the AI Screener We’re releasing a behind-the-scenes look at the Spectre AI Screener - narrated directly by the founder. This isn’t just a feature preview. It’s a dive into the thinking, structure, and direction behind what we’re building: - Multi-Tabs for seamless charting - Dexscan Orders directly integrated - AI Agents analyzing in real time - X Bubbles to track social flows - And a pipeline of utilities still unrevealed We're not waiting for BETA to optimize - core structural changes are already underway to strengthen the entire system before public release. A day mode is underway too. This is Spectre AI's next leap - built from conviction, precision, and relentless iteration. Not everything is covered in here - swaps, widgets, watchlists and more will be in the next thorough Voice-Over. It will be covered in the upcoming week. #aidex $SPECT

Sound ON/Founder Voice-Over: Inside the AI Screener We’re releasing a behind-the-scenes look at the Spectre AI Screener - narrated directly by the founder. This isn’t just a feature preview. It’s a dive into the thinking, structure, and direction behind what we’re building: - Multi-Tabs for seamless charting - Dexscan Orders directly integrated - AI Agents analyzing in real time - X Bubbles to track social flows - And a pipeline of utilities still unrevealed We're not waiting for BETA to optimize - core structural changes are already underway to strengthen the entire system before public release. A day mode is underway too. This is Spectre AI's next leap - built from conviction, precision, and relentless iteration. Not everything is covered in here - swaps, widgets, watchlists and more will be in the next thorough Voice-Over. It will be covered in the upcoming week. #aidex $SPECT

SPECTRE AI

23,531 views • 1 year ago

The AI gaming question I am most interested in is not whether AI can make better assets. It is whether video models can eventually become part of interactive worlds. PixVerse is exploring that question through PixVerse Game, combining real-time generative video, AI Agents, game mechanics, and user-defined worlds in an early technical experiment. My starting point for that question is here: I do not think AI gaming is solved, and this is not a finished platform. But if video models become more controllable and responsive, what role could they play inside future games? If the question interests you enough to test the idea yourself, share your take below. A few invitation codes are available, with priority going to the first thoughtful replies.

The AI gaming question I am most interested in is not whether AI can make better assets. It is whether video models can eventually become part of interactive worlds. PixVerse is exploring that question through PixVerse Game, combining real-time generative video, AI Agents, game mechanics, and user-defined worlds in an early technical experiment. My starting point for that question is here: I do not think AI gaming is solved, and this is not a finished platform. But if video models become more controllable and responsive, what role could they play inside future games? If the question interests you enough to test the idea yourself, share your take below. A few invitation codes are available, with priority going to the first thoughtful replies.

Deborah

77,604 views • 1 month ago

How many AI agents work at your company? We now have over 3,258 agents working alongside 1,300 humans. The crazy part is these agents were created by EVERY EMPLOYEE at our company... sales reps, marketers, customer support, product, eng. Literally EVERYONE. BUT I'm most surprised by the adoption and value that MANAGERS are getting from agents. I used to think that every IC would become a manager of agents. Now I think that managers will very likely manage WAY more agents than their ICs combined. And managers' agents will manage their ICs' agents - overseeing them for human-in-the-loop interactions. When creating agents, we use 100% context from all of your activity, files edited, tasks and projects worked on, hierarchy, skills, and role information. We build a user-based context model to make agents as relatable as possible to the specific human that we're building for. This means they truly understand the nuances of the work and what "great" looks like - because great is very much in the eye of the beholder. Great is by definition, subjective. This is also why the human ENGAGEMENT loops are SO vital to agent value. The iteration AFTER the agent is onboarded is where the MAGIC happens. This is just like a manager managing an IC in real life... you're giving feedback. In this case, though, agents learn INSTANTLY, and they retain the knowledge perfectly and indefinitely. Even though I've been pushing AI for years now to everyone in our company, this was the first time we had truly end-to-end AI adoption and retention. This kind of AI adoption is wild. But the value we're realizing is truly INSANE. Super Agents outnumber our humans nearly 3 to 1. What if you could 3X your workforce overnight? Watch this video to see how 👇

How many AI agents work at your company? We now have over 3,258 agents working alongside 1,300 humans. The crazy part is these agents were created by EVERY EMPLOYEE at our company... sales reps, marketers, customer support, product, eng. Literally EVERYONE. BUT I'm most surprised by the adoption and value that MANAGERS are getting from agents. I used to think that every IC would become a manager of agents. Now I think that managers will very likely manage WAY more agents than their ICs combined. And managers' agents will manage their ICs' agents - overseeing them for human-in-the-loop interactions. When creating agents, we use 100% context from all of your activity, files edited, tasks and projects worked on, hierarchy, skills, and role information. We build a user-based context model to make agents as relatable as possible to the specific human that we're building for. This means they truly understand the nuances of the work and what "great" looks like - because great is very much in the eye of the beholder. Great is by definition, subjective. This is also why the human ENGAGEMENT loops are SO vital to agent value. The iteration AFTER the agent is onboarded is where the MAGIC happens. This is just like a manager managing an IC in real life... you're giving feedback. In this case, though, agents learn INSTANTLY, and they retain the knowledge perfectly and indefinitely. Even though I've been pushing AI for years now to everyone in our company, this was the first time we had truly end-to-end AI adoption and retention. This kind of AI adoption is wild. But the value we're realizing is truly INSANE. Super Agents outnumber our humans nearly 3 to 1. What if you could 3X your workforce overnight? Watch this video to see how 👇

Zeb Evans

425,244 views • 7 months ago

Blender tutorials are about to age very badly. In this demo, Kimi K3 gets two character references and a plain-English request. Minutes later, it builds the first version inside Blender. The user asks for a better horn, Kimi edits the same scene, and the character starts moving. This is not another AI tool generating one lucky image. It is working with an actual editable 3D project, then changing the geometry after feedback instead of restarting from zero. It still needs human direction, because the first result is rarely the good one. But blockouts, repetitive edits, Blender Python, and basic iteration are becoming a conversation rather than hours of clicking through menus. Kimi K3 + Blender MCP is basically turning 3D modeling into: describe it, inspect it, complain about it, and let the AI fix it.

Blender tutorials are about to age very badly. In this demo, Kimi K3 gets two character references and a plain-English request. Minutes later, it builds the first version inside Blender. The user asks for a better horn, Kimi edits the same scene, and the character starts moving. This is not another AI tool generating one lucky image. It is working with an actual editable 3D project, then changing the geometry after feedback instead of restarting from zero. It still needs human direction, because the first result is rarely the good one. But blockouts, repetitive edits, Blender Python, and basic iteration are becoming a conversation rather than hours of clicking through menus. Kimi K3 + Blender MCP is basically turning 3D modeling into: describe it, inspect it, complain about it, and let the AI fix it.

Rina

19,092 views • 1 month ago

some advice for designing tools in the age of ai: 1. embrace ambiguity — ai tools work best when you're comfortable with not having full control. leave room for emergence and surprise. 2. focus on the human-ai collaboration model — design the interaction patterns, not just the outputs. what's the back-and-forth like? how do humans guide the ai? 3. design for iteration, not perfection — ai outputs are rarely perfect the first time. design systems that make refinement natural and easy. 4. think in systems, not features — ai changes fast, so build adaptable foundations rather than specific implementations that will break. 5. protect against the uncanny valley — either make it obviously ai-generated or indistinguishable from human work. the middle ground feels weird. 6. consider the invisible costs — ai has computational, environmental, and social costs that aren't immediately visible but matter. 7. design for augmentation, not replacement — the best ai tools make humans more capable rather than trying to replace them. 8. build in feedback loops — systems that learn from user interactions get better over time. the thing is, we're all figuring this out together. the design patterns for ai are still being invented. it's like the early web days — exciting, messy, and full of possibilities.

some advice for designing tools in the age of ai: 1. embrace ambiguity — ai tools work best when you're comfortable with not having full control. leave room for emergence and surprise. 2. focus on the human-ai collaboration model — design the interaction patterns, not just the outputs. what's the back-and-forth like? how do humans guide the ai? 3. design for iteration, not perfection — ai outputs are rarely perfect the first time. design systems that make refinement natural and easy. 4. think in systems, not features — ai changes fast, so build adaptable foundations rather than specific implementations that will break. 5. protect against the uncanny valley — either make it obviously ai-generated or indistinguishable from human work. the middle ground feels weird. 6. consider the invisible costs — ai has computational, environmental, and social costs that aren't immediately visible but matter. 7. design for augmentation, not replacement — the best ai tools make humans more capable rather than trying to replace them. 8. build in feedback loops — systems that learn from user interactions get better over time. the thing is, we're all figuring this out together. the design patterns for ai are still being invented. it's like the early web days — exciting, messy, and full of possibilities.

Ryo Lu

30,781 views • 1 year ago