{"type":"rich","version":"1.0","provider_name":"Transistor","provider_url":"https://transistor.fm","author_name":"AI News Today | Julian Goldie Podcast","title":"New Grok Build is ABSURD!","html":"","width":"100%","height":180,"duration":578,"description":"Grok Build Is Finally Good: 8-Agent CLI, X Search, and One-Sentence Apps (Plus Free with X Premium)\nThe video argues that Grok Build has improved dramatically and is now a surprisingly capable tool, especially for people who already have a Twitter/X subscription since it provides image generation, a CLI coding agent, and no apparent token limits. The creator shows how Grok Build can be used inside an agent setup (e.g., Hermes/OpenClora), including X/Twitter search and web search, and explains its workflow of planning first, splitting work across up to eight parallel sub-agents, self-checking, and iterating before shipping. Examples include building open-world games, simple websites, and an automated “video director” agent that generates fully narrated AI videos with avatars and B-roll. Benchmarks and side-by-side comparisons (Goldie Bench and a Voxel Racer test) suggest Grok Build is solid but generally not as strong as Claude Opus 4.8, and it now lists a 256K context window. The episode ends by promoting the AI Profit Boardroom and its agent operating system, tools, trainings, and coaching calls.\n00:00 Grok Build Surprise\n01:03 What Grok Build Is\n01:58 Games Built Fast\n02:20 Speed and Agents\n02:56 X Search Advantage\n03:31 How the CLI Works\n04:15 Best Way to Use It\n04:58 Websites and Video Creation\n06:08 Benchmark Results\n07:12 Voxel Game Comparison\n07:55 Context Window Update\n08:19 Agent OS and Boardroom Pitch\n09:35 Wrap Up","thumbnail_url":"https://img.transistorcdn.com/Tp0JSUlyiAe2ran44b13cjk4ImQYS1QEMKCAa9Q7go0/rs:fill:0:0:1/w:400/h:400/q:60/mb:500000/aHR0cHM6Ly9pbWct/dXBsb2FkLXByb2R1/Y3Rpb24udHJhbnNp/c3Rvci5mbS9lNTYz/NGQxMTU2ZTM1MzU5/MDNiYjcwZjJmYTY5/ODJmOS5qcGc.webp","thumbnail_width":300,"thumbnail_height":300}