martin_casado's banner

martin_casado

@martin_casado • 82,920 subscribers

GP @ a16z Nostalgic for the future

Shorts

Quick update. 5.3 Codex is a beast (!!): - Finally got permissions/policy in a good spot. Full support for distributed world building with per-user permissions on levels, items, NPCs - Multi-level / portals working - Deployment working - Now with extra sheep! (built with Cursor )

Quick update. 5.3 Codex is a beast (!!): - Finally got permissions/policy in a good spot. Full support for distributed world building with per-user permissions on levels, items, NPCs - Multi-level / portals working - Deployment working - Now with extra sheep! (built with Cursor )

46,828 Aufrufe

Super cool scene with flythrough. Really cool to see a 3D open world generated scene that is more than just a skybox with limited parallax

Super cool scene with flythrough. Really cool to see a 3D open world generated scene that is more than just a skybox with limited parallax

56,238 Aufrufe

AI is so trippy. This entire scene, including the exterior, was created from a single image of the ovens in the interior. This is a multi-view reconstruction, not a video. No v2v used. Prompt image in reply.

AI is so trippy. This entire scene, including the exterior, was created from a single image of the ovens in the interior. This is a multi-view reconstruction, not a video. No v2v used. Prompt image in reply.

54,363 Aufrufe

Basic combat system working and continuous tile animations. I think I have most of the pieces in place except for quests, and lots of level design to do.

Basic combat system working and continuous tile animations. I think I have most of the pieces in place except for quests, and lots of level design to do.

19,610 Aufrufe

Vibe coding in three.js is way better when the graphics don't look like total shit.

Vibe coding in three.js is way better when the graphics don't look like total shit.

45,422 Aufrufe

image -> 3D scene -> camera path -> wan1.3 v2v Pretty insane results. Playing with reconstruction now ...

image -> 3D scene -> camera path -> wan1.3 v2v Pretty insane results. Playing with reconstruction now ...

28,525 Aufrufe

Generative 3D scene workflows coming along. Took 11 hours to render 🤪

Generative 3D scene workflows coming along. Took 11 hours to render 🤪

26,031 Aufrufe

Quick Friday update. Support for complex NPC to NPC interactions. NPCs have relationship memory, internal state, dispositions, environment awareness, hidden agendas, ability to give / receive items, etc. Next, will have NPCs become players and take on quests etc.

Quick Friday update. Support for complex NPC to NPC interactions. NPCs have relationship memory, internal state, dispositions, environment awareness, hidden agendas, ability to give / receive items, etc. Next, will have NPCs become players and take on quests etc.

10,819 Aufrufe

Cozy sleeping spacepod. The geometry of the room is very interesting, particularly the rounded front of the room which is not part of the prompt image. Happy weekend everyone. Prompt image in reply ...

Cozy sleeping spacepod. The geometry of the room is very interesting, particularly the rounded front of the room which is not part of the prompt image. Happy weekend everyone. Prompt image in reply ...

17,268 Aufrufe

Another image -> 3D -> v2v (wan 2.1) render. The quality you can get is awesome. And because it's 3D mid workflow you can do arbitrary camera position / movement etc. Still 10 hour rendering time 🤡

Another image -> 3D -> v2v (wan 2.1) render. The quality you can get is awesome. And because it's 3D mid workflow you can do arbitrary camera position / movement etc. Still 10 hour rendering time 🤡

15,194 Aufrufe

I threw a couple of Mac Baconai's images in the new WorldLabs model. The scene completion is pretty darn good.

I threw a couple of Mac Baconai's images in the new WorldLabs model. The scene completion is pretty darn good.

20,678 Aufrufe

Great example of scene expansion. This was constructed entirely from the picture in the reply. Great parallax in the windows.

Great example of scene expansion. This was constructed entirely from the picture in the reply. Great parallax in the windows.

14,051 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

some of these 3D AI generations are getting out of hand. this is a stupidly large splat, it's not real time video.

some of these 3D AI generations are getting out of hand. this is a stupidly large splat, it's not real time video.

415,067 Aufrufe • vor 7 Monaten

OK, I'm really impressed. With Opus 4.6, Cursor and Convex I was able to get the following built in 4 hours: Fully persistent shared multiple player world with mutable object and NPC layer. Chat. Sprite editor. Map editor. Next, narrative logic for chat, inventory system, and combat framework.

OK, I'm really impressed. With Opus 4.6, Cursor and Convex I was able to get the following built in 4 hours: Fully persistent shared multiple player world with mutable object and NPC layer. Chat. Sprite editor. Map editor. Next, narrative logic for chat, inventory system, and combat framework.

165,835 Aufrufe • vor 3 Monaten

Total insanity. This is using an adaptive LOD scheme in sparksjs (not merged yet). The entire scene has 16million splats and this is real time navigation ... 😱😱

Total insanity. This is using an adaptive LOD scheme in sparksjs (not merged yet). The entire scene has 16million splats and this is real time navigation ... 😱😱

317,193 Aufrufe • vor 8 Monaten

The iteration speed with AI is just insane. Brittani Natali made the start scene in just a few hours. The total splat count for all these scenes together is around 400m or so. There is a *lot* to explore. Hosted version up at:

The iteration speed with AI is just insane. Brittani Natali made the start scene in just a few hours. The total splat count for all these scenes together is around 400m or so. There is a lot to explore. Hosted version up at:

120,236 Aufrufe • vor 5 Monaten

Some of the large scene's being generated are really unbelievable. This is a generative 3D splat, not 2D video. The full scene is likely over 50million splats. You can see some stitching inconsistencies from the final v2v pass. Also nice music :)

Some of the large scene's being generated are really unbelievable. This is a generative 3D splat, not 2D video. The full scene is likely over 50million splats. You can see some stitching inconsistencies from the final v2v pass. Also nice music :)

137,984 Aufrufe • vor 8 Monaten

Most angles have nice consistency in this AI generated 3D room. Objects have pretty OK geometry. Some ghosting. But regardless the pace of these models is just incredible.

Most angles have nice consistency in this AI generated 3D room. Objects have pretty OK geometry. Some ghosting. But regardless the pace of these models is just incredible.

117,786 Aufrufe • vor 9 Monaten

I just can't get enough of this 40million splat scene created by Vincent Woo. Rendered with spark ...

I just can't get enough of this 40million splat scene created by Vincent Woo. Rendered with spark ...

50,337 Aufrufe • vor 3 Monaten

Multi-splat object LOD test in sparkjs. 7 AI generated worlds, 12.5m total splats. Pretty consistent 120 fps on my laptop.

Multi-splat object LOD test in sparkjs. 7 AI generated worlds, 12.5m total splats. Pretty consistent 120 fps on my laptop.

95,404 Aufrufe • vor 7 Monaten

AI 3D workflows for web are getting scary good. Here is a demo with a large splat background (+LoD), mesh character, spatial audio, VR support, and lighting. I was able to create it in a few hours (repo in reply). built with Three.js spark World Labs Audio on :)

AI 3D workflows for web are getting scary good. Here is a demo with a large splat background (+LoD), mesh character, spatial audio, VR support, and lighting. I was able to create it in a few hours (repo in reply). built with Three.js spark World Labs Audio on :)

82,439 Aufrufe • vor 6 Monaten

The low poly meshes I created are much, much faster with physics / collision than those directly from the splat to mesh translation. Amazing. I now have an end to end workflow for generated 3D worlds with usable collisions. Works great in VR and mobile too ...

The low poly meshes I created are much, much faster with physics / collision than those directly from the splat to mesh translation. Amazing. I now have an end to end workflow for generated 3D worlds with usable collisions. Works great in VR and mobile too ...

54,452 Aufrufe • vor 4 Monaten

Working on a tool to generate low poly collision meshes from gaussian splats. It uses marching cubes and quadric error metric (QEM) decimation. Results are pretty good, it can get a usable sub 10Mb collision mesh from a 50Mb splat.

Working on a tool to generate low poly collision meshes from gaussian splats. It uses marching cubes and quadric error metric (QEM) decimation. Results are pretty good, it can get a usable sub 10Mb collision mesh from a 50Mb splat.

49,759 Aufrufe • vor 4 Monaten

You can pet the dog now ... All engine elements done, quests, AI NPCs, combat, items, multiplayer, portals, dynamic layers, multi-tilesets, interactive objects. Now it's just level/NPC design, performance/bug testing, and stats tweaking. (built with Convex and Cursor)

You can pet the dog now ... All engine elements done, quests, AI NPCs, combat, items, multiplayer, portals, dynamic layers, multi-tilesets, interactive objects. Now it's just level/NPC design, performance/bug testing, and stats tweaking. (built with Convex and Cursor)

41,796 Aufrufe • vor 3 Monaten

Update (8 hours development time): Built item layer, object interactions, multi-world / portal. Full live world/item/sprite/NPC editing. World is fully persistent with back-end loop managing NPCs etc. World is now fully buildable live, so you can edit as you go without requiring any restart (if you're an admin). All mutability of levels is reactive and updates multi-player. Multiplayer now smoother with movement prediction. Importantly, you can hang with the sleeping dog and cat. Next up, splash screens for interaction / combat. Built using Cursor and Convex primarily with 5.2-Codex and Opus 4.6.

Update (8 hours development time): Built item layer, object interactions, multi-world / portal. Full live world/item/sprite/NPC editing. World is fully persistent with back-end loop managing NPCs etc. World is now fully buildable live, so you can edit as you go without requiring any restart (if you're an admin). All mutability of levels is reactive and updates multi-player. Multiplayer now smoother with movement prediction. Importantly, you can hang with the sleeping dog and cat. Next up, splash screens for interaction / combat. Built using Cursor and Convex primarily with 5.2-Codex and Opus 4.6.

41,381 Aufrufe • vor 3 Monaten

AI aided 3D world building is getting ridiculous. This is a large, multi-room gaussian splat (~6million splats). Has streaming, LoD, character support, spatial audio, physics, collisions, VR support, and streaming video. Audio on ... Great place to watch a movie :)

AI aided 3D world building is getting ridiculous. This is a large, multi-room gaussian splat (~6million splats). Has streaming, LoD, character support, spatial audio, physics, collisions, VR support, and streaming video. Audio on ... Great place to watch a movie :)

47,482 Aufrufe • vor 4 Monaten

Collisions are finally coming! Automatic splat to conversion is tough given how messy AI generated splats are. So I created a simple editor that does down sampling, opacity filtering and then basic marching cubes and the results are pretty OK! 🥳

Collisions are finally coming! Automatic splat to conversion is tough given how messy AI generated splats are. So I created a simple editor that does down sampling, opacity filtering and then basic marching cubes and the results are pretty OK! 🥳

46,726 Aufrufe • vor 5 Monaten

I call it meditation coding. It's the opposite of vibe coding where you spend a ton of time meticulously hand coding something useless, when using AI or an app would have been way easier and faster.

I call it meditation coding. It's the opposite of vibe coding where you spend a ton of time meticulously hand coding something useless, when using AI or an app would have been way easier and faster.

98,528 Aufrufe • vor 1 Jahr

Another ridiculously large AI generated 3D scene. Tons of detail, and nice stereoscopic audio (turn on sound if you can). Works well in VR (although some quant effects from Spark near the end). Hosted version in reply.

Another ridiculously large AI generated 3D scene. Tons of detail, and nice stereoscopic audio (turn on sound if you can). Works well in VR (although some quant effects from Spark near the end). Hosted version in reply.

47,714 Aufrufe • vor 5 Monaten

Generative image to scene is getting incredibly good. This is pure 3D without v2v generated from the image in the reply ...

Generative image to scene is getting incredibly good. This is pure 3D without v2v generated from the image in the reply ...

57,089 Aufrufe • vor 9 Monaten

!! I really cannot believe this works. Large splat scene in VR on a Quest 3 using sparkjs. LoD budget is 500k splats. Runs about 30 fps. Full scene is ~6million splats. It's janky (sorry about the controls) But still, just insane this works at all ... (Metaverse is coming!)

!! I really cannot believe this works. Large splat scene in VR on a Quest 3 using sparkjs. LoD budget is 500k splats. Runs about 30 fps. Full scene is ~6million splats. It's janky (sorry about the controls) But still, just insane this works at all ... (Metaverse is coming!)

42,116 Aufrufe • vor 7 Monaten

Wild. Single shot text to 3D scene (gaussian splat) with no editing. This is approaching a usable scene asset for a game / 3D application.

Wild. Single shot text to 3D scene (gaussian splat) with no editing. This is approaching a usable scene asset for a game / 3D application.

56,136 Aufrufe • vor 10 Monaten