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.

Now I made it work. Let’s see the movement! Grok 4.6 built a running exploded view of a procedurally generated watch. SpaceXAI Grok 4.6 watch.

Dan

14,108 subscribers

832,804 views • 20 days ago •via X (Twitter)

Science & Technology

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

Grok 4.6 built this watch from scratch. Then took it apart. No scan. No downloaded model. Crystal, bezel, movement, date, crown, bracelet, built, previewed, and corrected until twelve sat at twelve. This is Grok 4.6.

Grok 4.6 built this watch from scratch. Then took it apart. No scan. No downloaded model. Crystal, bezel, movement, date, crown, bracelet, built, previewed, and corrected until twelve sat at twelve. This is Grok 4.6.

Dan

32,096 views • 21 days ago

ex-Apple engineer gave Grok 4.6 two jobs inside Cursor, went to sleep, and opened the results live the next morning: no babysitting, no checking every generation, just a model left running on real work for hours. • 00:42 - reveal the website Grok redesigned overnight • 23:21 - go from a voice prompt to a full software stack • 45:43 - inspect the generated code + architecture • 01:21:26 - Grok 4.6 vs Opus 5 • 01:46:31 - live PR + Cloud Agent workflow Most coding demos test an AI for 5 minutes. This tests the thing that matters for agents: can the model keep working when you stop watching it? SpaceXAI built Grok 4.6 specifically around longer-running agent tasks, coding and more ambitious visual work. The endgame isn’t prompting faster. It’s giving an agent a job at night and reviewing finished work in the morning. Worth watching before you choose which model runs your overnight agent loops.

ex-Apple engineer gave Grok 4.6 two jobs inside Cursor, went to sleep, and opened the results live the next morning: no babysitting, no checking every generation, just a model left running on real work for hours. • 00:42 - reveal the website Grok redesigned overnight • 23:21 - go from a voice prompt to a full software stack • 45:43 - inspect the generated code + architecture • 01:21:26 - Grok 4.6 vs Opus 5 • 01:46:31 - live PR + Cloud Agent workflow Most coding demos test an AI for 5 minutes. This tests the thing that matters for agents: can the model keep working when you stop watching it? SpaceXAI built Grok 4.6 specifically around longer-running agent tasks, coding and more ambitious visual work. The endgame isn’t prompting faster. It’s giving an agent a job at night and reviewing finished work in the morning. Worth watching before you choose which model runs your overnight agent loops.

Morlex

5,195,146 views • 10 days ago

Wow, Grok 4.6 making 3D print files for the real world! I told Grok I needed a spacer to be able to use the phone charger in my Tesla. From nothing, Grok found the required measurements, created 3D print files, and even made it multicolor for customization. It's printing now

Wow, Grok 4.6 making 3D print files for the real world! I told Grok I needed a spacer to be able to use the phone charger in my Tesla. From nothing, Grok found the required measurements, created 3D print files, and even made it multicolor for customization. It's printing now

Dirty Tesla

310,630 views • 19 days ago

Everything is coming up Grok these days. Grok Grok Bot. Grok 4.6. And Grok Github (or as it's officially named: Cursor origin) I cover it all in this week's mini-ep of How I AI, where I walk through: - my absolute favorite feature of Grok Bot - why multi agent identity > single agent identity - whether or not Origin is a real github killer - grok 4.6 design sense + where it beats opus/sol I also get real honest about why I'm keeping my OpenClaw🦞 around... for now 🦞🆘

Everything is coming up Grok these days. Grok Grok Bot. Grok 4.6. And Grok Github (or as it's officially named: Cursor origin) I cover it all in this week's mini-ep of How I AI, where I walk through: - my absolute favorite feature of Grok Bot - why multi agent identity > single agent identity - whether or not Origin is a real github killer - grok 4.6 design sense + where it beats opus/sol I also get real honest about why I'm keeping my OpenClaw🦞 around... for now 🦞🆘

claire vo 🖤

53,893 views • 15 days ago

Grok 4.6 created a real object to solve a real world issue and I was able to 3D print it. Huge.

Grok 4.6 created a real object to solve a real world issue and I was able to 3D print it. Huge.

Dirty Tesla

81,138 views • 19 days ago

Grok 4.6 is looking really good for making 3D games! Here's a quick demo of a little cosy 3d game working right within Unity. Used Meshy for the character from a Grok Imagine 2.0 image. Just prompts. I didn't touch anything in Unity or Blender. Grok 4.6 just did everything via the Unity CLI. SFX and BGM with ElevenLabs.

Grok 4.6 is looking really good for making 3D games! Here's a quick demo of a little cosy 3d game working right within Unity. Used Meshy for the character from a Grok Imagine 2.0 image. Just prompts. I didn't touch anything in Unity or Blender. Grok 4.6 just did everything via the Unity CLI. SFX and BGM with ElevenLabs.

Chong-U

225,787 views • 20 days ago

Woohoo! Thanks to my Game Developer Grok bot’s help, the game I’ve been building with Grok 4.6 now works great! My bot play-tested the game & pushed 10 PRs from Grok 4.6 through Cursor to fix bugs & make it run super smoothly! They're now working to design & build new levels😊

Woohoo! Thanks to my Game Developer Grok bot’s help, the game I’ve been building with Grok 4.6 now works great! My bot play-tested the game & pushed 10 PRs from Grok 4.6 through Cursor to fix bugs & make it run super smoothly! They're now working to design & build new levels😊

Derya Unutmaz, MD

11,088 views • 17 days ago

Grok 4.6 seems like a big improvement I gave it, along with Claude Opus 5, GPT 5.6 Sol, and Qwen 3.8 Max, a banana. Then asked them to be CREATIVE and do object-integrated drawings Here's how they turned out Grok 4.6 thought the banana can be a tiny island 😆

Grok 4.6 seems like a big improvement I gave it, along with Claude Opus 5, GPT 5.6 Sol, and Qwen 3.8 Max, a banana. Then asked them to be CREATIVE and do object-integrated drawings Here's how they turned out Grok 4.6 thought the banana can be a tiny island 😆

Ann Nguyen

250,219 views • 20 days ago

SuperGrok users can now redeem a usage limit reset! You can go to Usage and hit the button so you can go all in with Grok 4.6. Thank you SpaceXAI, this usage reset is a very generous present.

SuperGrok users can now redeem a usage limit reset! You can go to Usage and hit the button so you can go all in with Grok 4.6. Thank you SpaceXAI, this usage reset is a very generous present.

Testlabor

19,961 views • 21 days ago

Would you watch a full AI-generated nature documentary if it was actually good? Made with Grok Imagine 🎬

Would you watch a full AI-generated nature documentary if it was actually good? Made with Grok Imagine 🎬

Alex Utopia

10,798 views • 4 months ago

Took Grok 4.6 out for a spin this weekend and built this horde shooter prototype on ThreeJS. It is a big step up from Grok 4.5! I've always wanted to try building a horde shooter game with thousands of enemies onscreen and used Grok 4.6 for pretty much everything (including UI, vfx, and even the sfx!), except the 3D models for the characters, which were made and rigged with Meshy. I'm also amazed that ThreeJS with WebGPU can really handle thousands of enemies onscreen at the same time, though I think I pushed the limit here with Depth of Field and some other fancy shaders. Very fun!

Took Grok 4.6 out for a spin this weekend and built this horde shooter prototype on ThreeJS. It is a big step up from Grok 4.5! I've always wanted to try building a horde shooter game with thousands of enemies onscreen and used Grok 4.6 for pretty much everything (including UI, vfx, and even the sfx!), except the 3D models for the characters, which were made and rigged with Meshy. I'm also amazed that ThreeJS with WebGPU can really handle thousands of enemies onscreen at the same time, though I think I pushed the limit here with Depth of Field and some other fancy shaders. Very fun!

Danny Limanseta

63,198 views • 17 days ago

Grok Bot is already here and it’s different While regular Grok chats and answers questions, this new agent from xAI gets its own virtual machine, plugs into your tools, and actually finishes the work. A developer built a full sci-fi platformer with Grok 4.6 and Three.js, then handed it straight to Grok Bot for gameplay feedback. No more just talking about ideas. Start shipping them. Grok, Elon Musk Writer: Sol

Grok Bot is already here and it’s different While regular Grok chats and answers questions, this new agent from xAI gets its own virtual machine, plugs into your tools, and actually finishes the work. A developer built a full sci-fi platformer with Grok 4.6 and Three.js, then handed it straight to Grok Bot for gameplay feedback. No more just talking about ideas. Start shipping them. Grok, Elon Musk Writer: Sol

Mario Nawfal

59,768 views • 15 days ago

GLM 5.3 Max vs Fable 5 vs Qwen 3.8 Max vs Grok 4.6 GLM 5.3 is a massive jump over 5.2. Still not beating Fable 5 for me, but it’s way closer now. And vs Grok 4.6? I’d take GLM 5.3 pretty easily on this test.

GLM 5.3 Max vs Fable 5 vs Qwen 3.8 Max vs Grok 4.6 GLM 5.3 is a massive jump over 5.2. Still not beating Fable 5 for me, but it’s way closer now. And vs Grok 4.6? I’d take GLM 5.3 pretty easily on this test.

OmedTheVibeCoder

16,648 views • 19 days ago

$SITUATION EXPLAINED: How did Grok catch up this fast? • Grok 4.5 scored 56 on the Artificial Analysis index five weeks ago. Grok 4.6 scores 61, one behind Fable, at the same price • SOTA on knowledge work and legal, beating Fable on GDPval and Sol on Harvey LAB by 13 points Anastasios Nikolas Angelopoulos: "Grok 4.6 is a Sol-quality model, at a fraction of the price. Grok $6 vs Sol $30" sof 𓋹: "Now that the frontier labs can't release anything because, ooh, scary, close to critical capability thresholds, forget Chinese companies, all of the American companies are racing to close the gap."$

SITUATION EXPLAINED: How did Grok catch up this fast? • Grok 4.5 scored 56 on the Artificial Analysis index five weeks ago. Grok 4.6 scores 61, one behind Fable, at the same price • SOTA on knowledge work and legal, beating Fable on GDPval and Sol on Harvey LAB by 13 points Anastasios Nikolas Angelopoulos: "Grok 4.6 is a Sol-quality model, at a fraction of the price. Grok $6 vs Sol $30" sof 𓋹: "Now that the frontier labs can't release anything because, ooh, scary, close to critical capability thresholds, forget Chinese companies, all of the American companies are racing to close the gap."

MTS

34,203 views • 21 days ago

Grok Build just made web dev feel like cheating. A user built a full website covering every war in the 20th century, and Grok automatically generated image and video placeholders throughout the entire thing. Only @Grok

Grok Build just made web dev feel like cheating. A user built a full website covering every war in the 20th century, and Grok automatically generated image and video placeholders throughout the entire thing. Only @Grok

Mario Nawfal

39,675 views • 3 months ago

Grok 4.3 just built this entire game with just a single prompt It has the fastest output token speed and outranks Claude Sonnet 4.6 Max on Artificial Analysis I built this using the xAI API in Kilo Code via the VS Code extension

Grok 4.3 just built this entire game with just a single prompt It has the fastest output token speed and outranks Claude Sonnet 4.6 Max on Artificial Analysis I built this using the xAI API in Kilo Code via the VS Code extension

X Freeze

1,253,043 views • 4 months ago

SpaceXAI senior engineer just released a 1-hour workshop on running Grok 4.6 in Cursor and handing the rest to a GrokBot team - from scratch: • 01:14 - cutting token spend • 04:27 - harness engineering • 09:38 - model amnesia • 30:12 - how to build agents right • 44:08 - demo time: the whole stack running live This 1-hour watch will change how you think about delegating work to agents Skip Netflix tonight, give it an hour, then read the article below on automating your whole routine

SpaceXAI senior engineer just released a 1-hour workshop on running Grok 4.6 in Cursor and handing the rest to a GrokBot team - from scratch: • 01:14 - cutting token spend • 04:27 - harness engineering • 09:38 - model amnesia • 30:12 - how to build agents right • 44:08 - demo time: the whole stack running live This 1-hour watch will change how you think about delegating work to agents Skip Netflix tonight, give it an hour, then read the article below on automating your whole routine

Carnage

176,232 views • 8 days ago

I created a classic Doom-like game with Grok 4.6 build on my iPhone! It’s crazy that I can play it on my iPhone & it fully works! It didn’t even take an hour to make with a couple of prompts. I’m now going to recruit some Grok bots to play it & keep improving it! Then we party!😅

I created a classic Doom-like game with Grok 4.6 build on my iPhone! It’s crazy that I can play it on my iPhone & it fully works! It didn’t even take an hour to make with a couple of prompts. I’m now going to recruit some Grok bots to play it & keep improving it! Then we party!😅

Derya Unutmaz, MD

12,740 views • 17 days ago

SpaceXAI engineer (poteto): "The real mistake people make with GrokBot right now - building teams without trusting one bot’s output. So I built Chief of Staff bot that manages/reviews the other 10-20 agents " In 1-hour workshop SpaceXAI engineer explained the rules for initially setting up a 24/7 Grok Bot team, also the entire foundation of agent engineering and how she got started in coding Everyone who builds agents or plans to should watch this - replaces $500 courses Watch this and build first Grok Bot agent - then read the article below to set up the entire Grok Bot architecture ↓

SpaceXAI engineer (poteto): "The real mistake people make with GrokBot right now - building teams without trusting one bot’s output. So I built Chief of Staff bot that manages/reviews the other 10-20 agents " In 1-hour workshop SpaceXAI engineer explained the rules for initially setting up a 24/7 Grok Bot team, also the entire foundation of agent engineering and how she got started in coding Everyone who builds agents or plans to should watch this - replaces $500 courses Watch this and build first Grok Bot agent - then read the article below to set up the entire Grok Bot architecture ↓

codila

58,690 views • 4 days ago

Grok Build didn’t give beta much time to get comfortable. More than 100 updates in about 3 months, and now it’s officially V1.0. It can plan before touching your code, split work between parallel agents, work across files, use skills and MCP tools, remember previous sessions, and keep jobs running in the background. A lot of V1.0 is about making all of that more reliable when you’re actually using it for real work. Grok Build basically speedran beta. Grok SpaceXAI / Writer: Annette, Grok Imagine Designer: Janné

Grok Build didn’t give beta much time to get comfortable. More than 100 updates in about 3 months, and now it’s officially V1.0. It can plan before touching your code, split work between parallel agents, work across files, use skills and MCP tools, remember previous sessions, and keep jobs running in the background. A lot of V1.0 is about making all of that more reliable when you’re actually using it for real work. Grok Build basically speedran beta. Grok SpaceXAI / Writer: Annette, Grok Imagine Designer: Janné

Mario Nawfal

30,711 views • 22 days ago