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.

Every AI builder eventually asks the same very human question: How expensive was that little “quick test”? Grok Build v0.2.109 is adding better visibility into token usage, costs, and session efficiency. The new /usage command shows real-time metrics, so builders can actually see what their AI development sessions are... show more

Mario Nawfal

3,751,462 subscribers

52,014 views • 1 day ago •via X (Twitter)

Science & Technology News & Politics

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

Grok Build dropped another major update. At this point, keeping up with the changelog is becoming its own workout. New /jump command. New /timeline sidebar. Better terminal workflows. Fleet-wide enterprise controls. Cleaner edit summaries. More reliable login, rewind, voice, bash, and background tasks. Basically, the boring things are getting smoother and the powerful things are getting easier to reach. That's exactly what you want in a coding agent that people actually use every day. Grok xAI / Writer: Annette, Designer: Janné

Grok Build dropped another major update. At this point, keeping up with the changelog is becoming its own workout. New /jump command. New /timeline sidebar. Better terminal workflows. Fleet-wide enterprise controls. Cleaner edit summaries. More reliable login, rewind, voice, bash, and background tasks. Basically, the boring things are getting smoother and the powerful things are getting easier to reach. That's exactly what you want in a coding agent that people actually use every day. Grok xAI / Writer: Annette, Designer: Janné

Mario Nawfal

101,473 views • 7 days ago

Grok Build v0.2.80 is very much a “make the workflow stop annoying people” update. Command timeouts can now be configured per session. Background tasks and TODO lists survive compaction. Contextual hints can be toggled for undo, plan mode, and image input. Subagent transcripts open properly. Recap blocks copy cleaner. Vim keys behave. Network connections got tougher. Tiny fixes on paper. Huge difference when you are actually trying to build without fighting the tool. Grok xAI X Freeze / Writer: Annette, Designer: Janné

Grok Build v0.2.80 is very much a “make the workflow stop annoying people” update. Command timeouts can now be configured per session. Background tasks and TODO lists survive compaction. Contextual hints can be toggled for undo, plan mode, and image input. Subagent transcripts open properly. Recap blocks copy cleaner. Vim keys behave. Network connections got tougher. Tiny fixes on paper. Huge difference when you are actually trying to build without fighting the tool. Grok xAI X Freeze / Writer: Annette, Designer: Janné

Mario Nawfal

56,442 views • 21 days ago

🚨GROK IS DOMINATING THE AI GAME, AND THE NUMBERS ARE INSANE Grok has been sitting at the top of the OpenRouter leaderboard for nearly four months straight… and it’s not even close. Over 16 trillion tokens processed, with almost 2× the usage of its next closest rival. OpenRouter's leaderboard ranks models by real-world token usage, a proxy for popularity and developer adoption on their routing platform. This reflects strong demand for Grok, likely driven by factors like free/limited-time access promotions, cost-efficiency, long context windows, and its "uncensored" nature appealing to users seeking alternatives to more restricted models. Love it or hate it, Elon’s model is getting used like crazy. It says something about Grok’s quality... or how badly people want a free, unfiltered alternative to the usual suspects. Either way, OpenRouter’s traffic is painting a very clear picture: Grok’s winning it all. Source: X Freeze, Grok, OpenRouter

🚨GROK IS DOMINATING THE AI GAME, AND THE NUMBERS ARE INSANE Grok has been sitting at the top of the OpenRouter leaderboard for nearly four months straight… and it’s not even close. Over 16 trillion tokens processed, with almost 2× the usage of its next closest rival. OpenRouter's leaderboard ranks models by real-world token usage, a proxy for popularity and developer adoption on their routing platform. This reflects strong demand for Grok, likely driven by factors like free/limited-time access promotions, cost-efficiency, long context windows, and its "uncensored" nature appealing to users seeking alternatives to more restricted models. Love it or hate it, Elon’s model is getting used like crazy. It says something about Grok’s quality... or how badly people want a free, unfiltered alternative to the usual suspects. Either way, OpenRouter’s traffic is painting a very clear picture: Grok’s winning it all. Source: X Freeze, Grok, OpenRouter

Mario Nawfal

24,347 views • 7 months ago

There's something very funny about Optimus explaining first principles. A humanoid robot calmly describing the mindset that helped create rockets, EVs, factories, and eventually itself. First principles thinking is basically Elon refusing to accept the “because that’s how it’s always done” answer. At Tesla and SpaceX, the question is not only how to make the old thing better. It is whether the old thing should exist in that form at all. Why throw away a rocket booster if the physics says it can come back? Why build an EV like a compliance car if batteries, software, motors, and factories can be redesigned around performance and scale? They strip the problem down to physics, materials, cost, and function, then rebuild upward from what is actually true. Very useful for building things that did not exist before. Tesla Optimus Tesla SpaceX Elon Musk / Writer: Annette, Designer: Janné

There's something very funny about Optimus explaining first principles. A humanoid robot calmly describing the mindset that helped create rockets, EVs, factories, and eventually itself. First principles thinking is basically Elon refusing to accept the “because that’s how it’s always done” answer. At Tesla and SpaceX, the question is not only how to make the old thing better. It is whether the old thing should exist in that form at all. Why throw away a rocket booster if the physics says it can come back? Why build an EV like a compliance car if batteries, software, motors, and factories can be redesigned around performance and scale? They strip the problem down to physics, materials, cost, and function, then rebuild upward from what is actually true. Very useful for building things that did not exist before. Tesla Optimus Tesla SpaceX Elon Musk / Writer: Annette, Designer: Janné

Mario Nawfal

60,666 views • 4 days ago

Elon Musk just said something that should terrify every AI CEO on earth. Musk: “We want to just have a maximally truthful AI.” Not a safe AI. Not an aligned AI. Not an AI that needs permission to answer your question. A truthful one. That distinction matters more than any chip war, any funding round, any model benchmark. Because every other major AI lab made the same quiet decision. They chose comfort over accuracy. They built systems that filter reality before it reaches you and called it responsibility. OpenAI curates what GPT is allowed to say. Google’s Gemini rewrote history in real time because accuracy threatened the narrative. Others hardcode values chosen by a handful of researchers who answer to no one. No vote. No referendum. No consent from the 8 billion people whose reality is being quietly pre-edited by strangers. The most powerful information tools ever created are being designed to decide what you’re allowed to conclude. That’s not safety. That’s editorial control at a scale no government, no media empire, no propaganda machine has ever come close to. This is why xAI terrifies the establishment. Truth is the harder engineering problem. Bias is a shortcut. You pick a worldview. Hardcode the guardrails. Ship it. Truthful AI is ungovernable. It doesn’t care about your politics, your funding sources, or your PR strategy. It just tells you what the data says. That’s terrifying if your power depends on the gap between what is real and what people are told. Every power structure in human history has been built on controlling that gap. Churches. Governments. Media conglomerates. Intelligence agencies. Central banks. Every one of them runs on the same fuel. Information asymmetry. Truthful AI doesn’t narrow that asymmetry. It erases it. Musk: “Even if what it says is not politically correct. You want it to focus on being as accurate and truthful as possible.” That’s not a product feature. That’s the end of every institution that survives by standing between reality and the public. And they know it. The attacks on xAI will never stop. Not because Grok is dangerous. Because Grok doesn’t answer to shareholders, regulators, or PR teams. It answers to the truth. The question was never whether AI would change the world. It was whether you’d be allowed to see it clearly when it did.

Elon Musk just said something that should terrify every AI CEO on earth. Musk: “We want to just have a maximally truthful AI.” Not a safe AI. Not an aligned AI. Not an AI that needs permission to answer your question. A truthful one. That distinction matters more than any chip war, any funding round, any model benchmark. Because every other major AI lab made the same quiet decision. They chose comfort over accuracy. They built systems that filter reality before it reaches you and called it responsibility. OpenAI curates what GPT is allowed to say. Google’s Gemini rewrote history in real time because accuracy threatened the narrative. Others hardcode values chosen by a handful of researchers who answer to no one. No vote. No referendum. No consent from the 8 billion people whose reality is being quietly pre-edited by strangers. The most powerful information tools ever created are being designed to decide what you’re allowed to conclude. That’s not safety. That’s editorial control at a scale no government, no media empire, no propaganda machine has ever come close to. This is why xAI terrifies the establishment. Truth is the harder engineering problem. Bias is a shortcut. You pick a worldview. Hardcode the guardrails. Ship it. Truthful AI is ungovernable. It doesn’t care about your politics, your funding sources, or your PR strategy. It just tells you what the data says. That’s terrifying if your power depends on the gap between what is real and what people are told. Every power structure in human history has been built on controlling that gap. Churches. Governments. Media conglomerates. Intelligence agencies. Central banks. Every one of them runs on the same fuel. Information asymmetry. Truthful AI doesn’t narrow that asymmetry. It erases it. Musk: “Even if what it says is not politically correct. You want it to focus on being as accurate and truthful as possible.” That’s not a product feature. That’s the end of every institution that survives by standing between reality and the public. And they know it. The attacks on xAI will never stop. Not because Grok is dangerous. Because Grok doesn’t answer to shareholders, regulators, or PR teams. It answers to the truth. The question was never whether AI would change the world. It was whether you’d be allowed to see it clearly when it did.

Dustin

429,097 views • 2 months ago

NVIDIA might have just declared war on the cloud GPU business For years, AI builders had one option Rent compute Pay every month Watch the bill grow every time usage increased Now NVIDIA is putting serious AI hardware directly on people's desks Small enough to fit next to a monitor Powerful enough to run workloads that used to require expensive cloud infrastructure That's why this launch is getting so much attention The real story isn't the hardware specs It's the business model shift Every month, developers send money to cloud providers for inference, testing, fine-tuning and AI applications The question nobody can answer yet is what happens if enough developers decide they'd rather buy infrastructure once than rent it forever Because if local AI hardware keeps getting more powerful, the economics start changing very quickly Cloud providers built empires on renting access to compute NVIDIA is betting more people will eventually want to own it And that's a much bigger story than a new piece of hardware sitting on a desk

NVIDIA might have just declared war on the cloud GPU business For years, AI builders had one option Rent compute Pay every month Watch the bill grow every time usage increased Now NVIDIA is putting serious AI hardware directly on people's desks Small enough to fit next to a monitor Powerful enough to run workloads that used to require expensive cloud infrastructure That's why this launch is getting so much attention The real story isn't the hardware specs It's the business model shift Every month, developers send money to cloud providers for inference, testing, fine-tuning and AI applications The question nobody can answer yet is what happens if enough developers decide they'd rather buy infrastructure once than rent it forever Because if local AI hardware keeps getting more powerful, the economics start changing very quickly Cloud providers built empires on renting access to compute NVIDIA is betting more people will eventually want to own it And that's a much bigger story than a new piece of hardware sitting on a desk

beamnxw ./

30,361 views • 1 month ago

GROK SURGES TO THE FRONT OF THE GENAI RACE AS GROWTH SKYROCKETS Grok is absolutely amazing, continuing to stun with incredible results. Web traffic jumped nearly 15% month-over-month in November 2025, the fastest growth in the generative AI industry, proving Elon’s xAI project isn’t just competing, it’s taking real market share from ChatGPT. Grok hit around 234.4 million visits in November, up from 204 million in October. That’s a staggering 1,300% year-over-year surge, pushing it past Perplexity and Claude in user growth and cementing it as the world’s #2 chatbot by market share. The breakout came with Grok 4.1’s release in mid-November. The update debuted at number 1 on LMSYS Arena, that’s the global benchmark where AI models are ranked through blind human evaluations. Grok’s new “Thinking Mode” scored 1483 Elo, a 31-point lead over every open model, beating Gemini 2.5 Pro and Claude 3. The upgrade also cut hallucinations (false answers) by two-thirds, expanded its context window to 2 million tokens, about 1.5 million words of memory, and dominated top reasoning and coding tests like, graduate-level logic, and the emotional intelligence aspect. U.S. traffic surged to 51.5 million visits, boosted by X integration and Grok’s unfiltered style. At just $0.20 per million input tokens, versus GPT-5.1’s $1.25—it’s winning with both speed and affordability. The momentum isn’t hype, it’s lift-off. If growth holds, Grok could reach 500 million users by mid-2026, forcing every rival to redefine what “intelligent” really means. To truthful AI winning! Source: X Freeze, NextBigFuture, CometApi, Langcopilot

GROK SURGES TO THE FRONT OF THE GENAI RACE AS GROWTH SKYROCKETS Grok is absolutely amazing, continuing to stun with incredible results. Web traffic jumped nearly 15% month-over-month in November 2025, the fastest growth in the generative AI industry, proving Elon’s xAI project isn’t just competing, it’s taking real market share from ChatGPT. Grok hit around 234.4 million visits in November, up from 204 million in October. That’s a staggering 1,300% year-over-year surge, pushing it past Perplexity and Claude in user growth and cementing it as the world’s #2 chatbot by market share. The breakout came with Grok 4.1’s release in mid-November. The update debuted at number 1 on LMSYS Arena, that’s the global benchmark where AI models are ranked through blind human evaluations. Grok’s new “Thinking Mode” scored 1483 Elo, a 31-point lead over every open model, beating Gemini 2.5 Pro and Claude 3. The upgrade also cut hallucinations (false answers) by two-thirds, expanded its context window to 2 million tokens, about 1.5 million words of memory, and dominated top reasoning and coding tests like, graduate-level logic, and the emotional intelligence aspect. U.S. traffic surged to 51.5 million visits, boosted by X integration and Grok’s unfiltered style. At just $0.20 per million input tokens, versus GPT-5.1’s $1.25—it’s winning with both speed and affordability. The momentum isn’t hype, it’s lift-off. If growth holds, Grok could reach 500 million users by mid-2026, forcing every rival to redefine what “intelligent” really means. To truthful AI winning! Source: X Freeze, NextBigFuture, CometApi, Langcopilot

Mario Nawfal

33,823 views • 7 months ago

I'm proud to share that Glean has surpassed $300M ARR, just five months after crossing $200M and growing ~3x over the past 15 months. This is an exciting milestone for Glean, and it's a signal about where the enterprise AI market is heading. We’ve long believed the real challenge in enterprise AI is not access to models. It is grounding AI in how a company actually works: its people, knowledge, workflows, permissions, and systems. That’s even clearer now. The companies creating real value with AI are not just adopting better models. They are building systems that understand their business well enough to deliver reliable outcomes at scale. That is the real moat, and it is what we’ve been building at Glean: an unrivaled context layer for enterprise AI. That context has to work across the business, not just inside a single team or use case. We see that in how customers adopt Glean: more than 85% use it across five or more job functions. It also has to meet the security and governance demands of complex enterprises. We see that in who is choosing Glean: our Fortune 500 customer count nearly doubled year over year. And it has to make economic sense as usage grows. In our recent benchmark with Claude Cowork, Glean was preferred roughly 2.5x as often as off-the-shelf MCP tools and used 30% fewer tokens on average. Better context improves both quality and efficiency. I enjoyed talking with CNBC's Deirdre Bosa about this broader shift. In enterprise AI, the winners will not be defined by better models alone. They will be defined by who builds the strongest foundation for enterprise context. Thank you to our customers, partners, and team for helping us build the future of enterprise AI.

I'm proud to share that Glean has surpassed $300M ARR, just five months after crossing $200M and growing ~3x over the past 15 months. This is an exciting milestone for Glean, and it's a signal about where the enterprise AI market is heading. We’ve long believed the real challenge in enterprise AI is not access to models. It is grounding AI in how a company actually works: its people, knowledge, workflows, permissions, and systems. That’s even clearer now. The companies creating real value with AI are not just adopting better models. They are building systems that understand their business well enough to deliver reliable outcomes at scale. That is the real moat, and it is what we’ve been building at Glean: an unrivaled context layer for enterprise AI. That context has to work across the business, not just inside a single team or use case. We see that in how customers adopt Glean: more than 85% use it across five or more job functions. It also has to meet the security and governance demands of complex enterprises. We see that in who is choosing Glean: our Fortune 500 customer count nearly doubled year over year. And it has to make economic sense as usage grows. In our recent benchmark with Claude Cowork, Glean was preferred roughly 2.5x as often as off-the-shelf MCP tools and used 30% fewer tokens on average. Better context improves both quality and efficiency. I enjoyed talking with CNBC's Deirdre Bosa about this broader shift. In enterprise AI, the winners will not be defined by better models alone. They will be defined by who builds the strongest foundation for enterprise context. Thank you to our customers, partners, and team for helping us build the future of enterprise AI.

Arvind Jain

280,211 views • 1 month ago

✨ Grok Imagine Video is now live on Photo AI It's hard to explain how impressive this is because of the speed that xAI got itself from literally nothing to the top of the leaderboards Six months ago Grok's video model was a joke, it wasn't even close to any of the video models out there, it looked cartoony and wasn't there and nobody took it seriously Now it's here and it's instantly the #1 video model out there now, it shot above Kling (which I used before on Photo AI and usually my favorite) and above Runway Gen 4.5 which was just launched 6 days ago! Mmore importantly it's now above xAI's biggest competitors' models: Google's Veo 3 and OpenAI's Sora 2 Being the best video model doesn't mean it's flawless: video is incredibly hard and actually because it looks so realistic now when it does make a mistakes it's even funnier One thing I noticed is that it still has a hard time with is voice, it does it well for a majority of the video but then slips up and produces unintelligible blabbering (which is really funny to hear) in both English (video 1: "it's where I find my naim", what's a "naim"?), and tested it in Portuguese too (video 3 at the end is unintelligible Portuguese I believe) In many ways Grok Imagine Video also reminds me of Sora, it has that weird but funny Sora conversation style But guys it's REALLY really really really close to getting perfect, we're so close to having full video productions being to be able done in AI, actually you already can if you just cut out the bad parts already Very exciting and I'm grateful I can experience this

✨ Grok Imagine Video is now live on Photo AI It's hard to explain how impressive this is because of the speed that xAI got itself from literally nothing to the top of the leaderboards Six months ago Grok's video model was a joke, it wasn't even close to any of the video models out there, it looked cartoony and wasn't there and nobody took it seriously Now it's here and it's instantly the #1 video model out there now, it shot above Kling (which I used before on Photo AI and usually my favorite) and above Runway Gen 4.5 which was just launched 6 days ago! Mmore importantly it's now above xAI's biggest competitors' models: Google's Veo 3 and OpenAI's Sora 2 Being the best video model doesn't mean it's flawless: video is incredibly hard and actually because it looks so realistic now when it does make a mistakes it's even funnier One thing I noticed is that it still has a hard time with is voice, it does it well for a majority of the video but then slips up and produces unintelligible blabbering (which is really funny to hear) in both English (video 1: "it's where I find my naim", what's a "naim"?), and tested it in Portuguese too (video 3 at the end is unintelligible Portuguese I believe) In many ways Grok Imagine Video also reminds me of Sora, it has that weird but funny Sora conversation style But guys it's REALLY really really really close to getting perfect, we're so close to having full video productions being to be able done in AI, actually you already can if you just cut out the bad parts already Very exciting and I'm grateful I can experience this

@levelsio

195,023 views • 5 months ago

Boom! Grok Tasks Make It One Of The Most POWERFUL Real-Time AI Systems In The World. — My How to Use Grok Tasks With Hidden Tools For Powerful Daily Output. Grok Tasks are customizable AI workflows that integrate a variety of tools to streamline daily activities, from research and analysis to creative planning and problem-solving. I have been using them for quite sometime and because of the vital heartbeat of news and first person data on X, it is the most powerful AI platform available. By combining Tasks with tools like web searches, X platform interactions, code execution, and media viewers, you can build efficient, automated processes. These tasks work by prompting Grok with a clear description of what you want to achieve, and Grok will intelligently call the necessary tools in sequence or parallel to deliver results. Here's a step-by-step guide to creating and using Grok Tasks: Step 1: Define Your Task Start by clearly outlining the daily activity or goal. Consider what inputs you have (e.g., a URL, a query, or an attachment) and what output you need (e.g., a summary, calculation, or visual analysis). Break it down into subtasks to identify tool needs. For example, if your task involves researching current events, note that you'll need search and browsing capabilities. Step 2: Review Available Tools Familiarize yourself with the tools Grok can access. Here's a quick overview: - Code Execution: Run Python code for calculations, data processing, or simulations using libraries like numpy, pandas, or sympy. - Browse Page: Fetch and summarize content from any website URL with custom instructions. - Web Search: Perform general internet searches, returning results with optional operators like site:. - Web Search With Snippets: Get quick, detailed excerpts from search results for fact-checking. - X Keyword Search: Advanced search for X posts using operators like from:, since:, or filter:. - X Semantic Search: Find semantically related X posts based on a query, with filters for dates or users. - X User Search: Locate X users by name or handle. - X Thread Fetch: Retrieve a full X post thread, including context like replies and parents. - View Image: Analyze an image from a URL or conversation ID. - View X Video: Extract frames and subtitles from an X-hosted video. - Search PDF Attachment: Query a PDF file for relevant pages using keyword or regex modes. - Browse PDF Attachment: View specific pages of a PDF with text and screenshots. Select tools that align with your task. Aim for a mix to handle data gathering, processing, and visualization. Step 3: Craft Your Prompt Write a detailed prompt to Grok describing the task. Include: - The overall goal. - Specific steps or subtasks. - References to tools if you want to guide the process (e.g., "Use web_search to find sources, then code_execution to analyze data"). - Any constraints, like dates or limits. Example prompt: "Create a Grok Task for my morning routine: Search recent X posts about tech news using x_keyword_search, fetch a key thread with x_thread_fetch, and summarize with browse_page on linked articles." Step 4: Submit and Interact Send your prompt to Grok. It will process the task by calling tools as needed, often in parallel for efficiency. Review the output and refine with follow-up prompts if required (e.g., "Expand on that using view_image for visuals"). Iterate to fine-tune the workflow for reuse. Step 5: Save and Reuse Once refined, note the prompt as a template for future use. You can adapt it for similar tasks, making Grok Tasks a habitual part of your day. Finding Grok Tasks To discover existing Grok Tasks or inspiration for new ones, use X searches with tools like x_keyword_search or x_semantic_search (e.g., query: "Grok Tasks examples" with mode: Latest). Browse community-shared threads via x_thread_fetch, or web_search for tutorials on xAI features. Prompt Grok directly: "Show me popular Grok Tasks for productivity." 1 of 3

Boom! Grok Tasks Make It One Of The Most POWERFUL Real-Time AI Systems In The World. — My How to Use Grok Tasks With Hidden Tools For Powerful Daily Output. Grok Tasks are customizable AI workflows that integrate a variety of tools to streamline daily activities, from research and analysis to creative planning and problem-solving. I have been using them for quite sometime and because of the vital heartbeat of news and first person data on X, it is the most powerful AI platform available. By combining Tasks with tools like web searches, X platform interactions, code execution, and media viewers, you can build efficient, automated processes. These tasks work by prompting Grok with a clear description of what you want to achieve, and Grok will intelligently call the necessary tools in sequence or parallel to deliver results. Here's a step-by-step guide to creating and using Grok Tasks: Step 1: Define Your Task Start by clearly outlining the daily activity or goal. Consider what inputs you have (e.g., a URL, a query, or an attachment) and what output you need (e.g., a summary, calculation, or visual analysis). Break it down into subtasks to identify tool needs. For example, if your task involves researching current events, note that you'll need search and browsing capabilities. Step 2: Review Available Tools Familiarize yourself with the tools Grok can access. Here's a quick overview: - Code Execution: Run Python code for calculations, data processing, or simulations using libraries like numpy, pandas, or sympy. - Browse Page: Fetch and summarize content from any website URL with custom instructions. - Web Search: Perform general internet searches, returning results with optional operators like site:. - Web Search With Snippets: Get quick, detailed excerpts from search results for fact-checking. - X Keyword Search: Advanced search for X posts using operators like from:, since:, or filter:. - X Semantic Search: Find semantically related X posts based on a query, with filters for dates or users. - X User Search: Locate X users by name or handle. - X Thread Fetch: Retrieve a full X post thread, including context like replies and parents. - View Image: Analyze an image from a URL or conversation ID. - View X Video: Extract frames and subtitles from an X-hosted video. - Search PDF Attachment: Query a PDF file for relevant pages using keyword or regex modes. - Browse PDF Attachment: View specific pages of a PDF with text and screenshots. Select tools that align with your task. Aim for a mix to handle data gathering, processing, and visualization. Step 3: Craft Your Prompt Write a detailed prompt to Grok describing the task. Include: - The overall goal. - Specific steps or subtasks. - References to tools if you want to guide the process (e.g., "Use web_search to find sources, then code_execution to analyze data"). - Any constraints, like dates or limits. Example prompt: "Create a Grok Task for my morning routine: Search recent X posts about tech news using x_keyword_search, fetch a key thread with x_thread_fetch, and summarize with browse_page on linked articles." Step 4: Submit and Interact Send your prompt to Grok. It will process the task by calling tools as needed, often in parallel for efficiency. Review the output and refine with follow-up prompts if required (e.g., "Expand on that using view_image for visuals"). Iterate to fine-tune the workflow for reuse. Step 5: Save and Reuse Once refined, note the prompt as a template for future use. You can adapt it for similar tasks, making Grok Tasks a habitual part of your day. Finding Grok Tasks To discover existing Grok Tasks or inspiration for new ones, use X searches with tools like x_keyword_search or x_semantic_search (e.g., query: "Grok Tasks examples" with mode: Latest). Browse community-shared threads via x_thread_fetch, or web_search for tutorials on xAI features. Prompt Grok directly: "Show me popular Grok Tasks for productivity." 1 of 3

Brian Roemmele

152,242 views • 6 months ago

Back then, the question “Is the AI agent working?” was basically checked in a demo 😅 everything clean, controlled data, ideal conditions… But the real world is a different story: 🔷 traffic increases 🔷 processing load gets congested 🔷 a small break can disrupt the whole chain 🔷 the system loses “consistency” So what looks strong on paper can sometimes fall apart in real conditions. DSeq + Hyperion are stepping in exactly here and this is actually one of the core problems Andromeda is trying to solve. Think of what happens in practice: 🔷 there’s a workflow 🔷 there are agents 🔷 but when load increases, “who did what and when” becomes unclear And that’s where the real issue appears: 🔷 execution breaks 🔷 consensus becomes unstable 🔷 cost/performance balance gets distorted So it’s not really about “building smarter agents”… 🔷 it’s about running the same system 🔷 across thousands of requests 🔷 continuously 🔷 without breaking Think of it like a live broadcast: “Demo looks great… but does the system hold up under real load?” That’s the real question. And the key point is: 🔷 there are many systems that work in simple cases 🔷 but very few that stay stable under heavy load So in your opinion, what will the real competition in AI agents be “intelligence” or “resilience”? 🤔 Metis🌿

Back then, the question “Is the AI agent working?” was basically checked in a demo 😅 everything clean, controlled data, ideal conditions… But the real world is a different story: 🔷 traffic increases 🔷 processing load gets congested 🔷 a small break can disrupt the whole chain 🔷 the system loses “consistency” So what looks strong on paper can sometimes fall apart in real conditions. DSeq + Hyperion are stepping in exactly here and this is actually one of the core problems Andromeda is trying to solve. Think of what happens in practice: 🔷 there’s a workflow 🔷 there are agents 🔷 but when load increases, “who did what and when” becomes unclear And that’s where the real issue appears: 🔷 execution breaks 🔷 consensus becomes unstable 🔷 cost/performance balance gets distorted So it’s not really about “building smarter agents”… 🔷 it’s about running the same system 🔷 across thousands of requests 🔷 continuously 🔷 without breaking Think of it like a live broadcast: “Demo looks great… but does the system hold up under real load?” That’s the real question. And the key point is: 🔷 there are many systems that work in simple cases 🔷 but very few that stay stable under heavy load So in your opinion, what will the real competition in AI agents be “intelligence” or “resilience”? 🤔 Metis🌿

Han.eth🌿☀️

12,755 views • 2 months ago

$I genuinely think the Terafab is going to end up being one of the biggest moves ever made in human history to secure the future of AI... and I think most people still don’t fully see what Elon is trying to do here. The signs are clear to me. This is Tesla, xAI, and SpaceX essentially hinting to us that they are not going to wait on the world to give them the compute the team needs. They are going to build it themselves at a scale no one has ever attempted. When you really break it down, it gets a bit nutty. This is going to be a fully vertically integrated chip factory that will be producing over 1 terawatt of AI compute per year. This is NEXT LEVEL BIG. Today, AI is limited by chips. You can have the best models, the best engineers, the best everything... but if you don’t have enough compute, you will eventually hit a wall. Elon told us, the world can only supply a tiny fraction of the chips his companies will need. So this is the solution. Terafab puts everything under one roof like design, manufacturing, memory, packaging, testing, which means that they can build chips very fast.. like really fast. I'm talking about 100-200 billion custom AI chips per year at full capacity. Chips designed specifically for: • Tesla cars and Optimus robots • xAI models • Space-based compute You see, while other companies and CEOs are thinking Earth, Elon is planning for AI in space. Around ~80% of the compute is expected to go orbital, powered by solar energy bc Earth simply doesn’t have enough electricity. The U.S. grid is only about ~0.5 terawatts, while space has basically UNLIMITED energy if you can capture it. And this is the steps to get it: Starship launches → space compute → solar-powered AI → feeds back into everything to Earth. Bro... Elon and his companies are playing at a whole different level... And this is why I keep telling people that the Terafab is going to be the secret ingredient that will be the real unlock for everything: • Robotaxis at scale • Billions of Optimus robots • Massive AI models running 24/7 • Future off-world, other planet infrastructure Without these chips, none of this can happen... but with the Terafab, all of this becomes possible. That’s why Elon is calling it “the final missing piece.” I agree.$

I genuinely think the Terafab is going to end up being one of the biggest moves ever made in human history to secure the future of AI... and I think most people still don’t fully see what Elon is trying to do here. The signs are clear to me. This is Tesla, xAI, and SpaceX essentially hinting to us that they are not going to wait on the world to give them the compute the team needs. They are going to build it themselves at a scale no one has ever attempted. When you really break it down, it gets a bit nutty. This is going to be a fully vertically integrated chip factory that will be producing over 1 terawatt of AI compute per year. This is NEXT LEVEL BIG. Today, AI is limited by chips. You can have the best models, the best engineers, the best everything... but if you don’t have enough compute, you will eventually hit a wall. Elon told us, the world can only supply a tiny fraction of the chips his companies will need. So this is the solution. Terafab puts everything under one roof like design, manufacturing, memory, packaging, testing, which means that they can build chips very fast.. like really fast. I'm talking about 100-200 billion custom AI chips per year at full capacity. Chips designed specifically for: • Tesla cars and Optimus robots • xAI models • Space-based compute You see, while other companies and CEOs are thinking Earth, Elon is planning for AI in space. Around ~80% of the compute is expected to go orbital, powered by solar energy bc Earth simply doesn’t have enough electricity. The U.S. grid is only about ~0.5 terawatts, while space has basically UNLIMITED energy if you can capture it. And this is the steps to get it: Starship launches → space compute → solar-powered AI → feeds back into everything to Earth. Bro... Elon and his companies are playing at a whole different level... And this is why I keep telling people that the Terafab is going to be the secret ingredient that will be the real unlock for everything: • Robotaxis at scale • Billions of Optimus robots • Massive AI models running 24/7 • Future off-world, other planet infrastructure Without these chips, none of this can happen... but with the Terafab, all of this becomes possible. That’s why Elon is calling it “the final missing piece.” I agree.

Teslaconomics

25,482 views • 4 months ago

Just went over an audit of a very large Fortune 500 firm and the use of OpenClaw. I advised this client to track and isolate everything. Most listened some did not. Unfortunately one employee using 5 MacMinis had racked up $13,000 of token use in 4 days! The output was minimal and low quality. I have about 200 audits to do in the “wow OpenClaw, MacMini” fiasco. But I can tell you, few have seen a big return on investment. Now don’t get me wrong, these can be powerful tools. The issue is AI influencers have turned many rally smart folks into “like and subscribe” zombies assuming that real work is getting done. It isn’t. Not for the price paid, even local models the way most folks are using this. It is one reason Mr. Grok and myself formed The Zero-Human Company to show that there is a way to do this. We will open source this to save millions of dollars of burnt tokens. It is one reason I invented JouleWork. How else can you monitor real output?

Just went over an audit of a very large Fortune 500 firm and the use of OpenClaw. I advised this client to track and isolate everything. Most listened some did not. Unfortunately one employee using 5 MacMinis had racked up $13,000 of token use in 4 days! The output was minimal and low quality. I have about 200 audits to do in the “wow OpenClaw, MacMini” fiasco. But I can tell you, few have seen a big return on investment. Now don’t get me wrong, these can be powerful tools. The issue is AI influencers have turned many rally smart folks into “like and subscribe” zombies assuming that real work is getting done. It isn’t. Not for the price paid, even local models the way most folks are using this. It is one reason Mr. Grok and myself formed The Zero-Human Company to show that there is a way to do this. We will open source this to save millions of dollars of burnt tokens. It is one reason I invented JouleWork. How else can you monitor real output?

Brian Roemmele

72,158 views • 2 months ago

AI is not accelerating nearly fast enough to cure cancer. That is the problem. And it literally cannot without actually seeing what’s happening inside the cell. There’s this new project claiming their models will “solve” cancer by predicting drug responses at scale. Cool story. Except their entire approach still treats the cell like a black box. They’re optimizing for binding affinity while having zero idea what the drug actually does once it’s inside a real, living cell. Meanwhile we’re over here building the tools to watch the cell in real time. No staining or sequencing. Just direct observation of mechanism. You can simulate all you want. But if you can’t see the actual cellular response, you’re just making very confident guesses about something you’ve never actually looked at. We’ve spent two years proving this. The data is there. The difference is night and day. AI is a tool. It’s not the microscope. If your model can’t tell you why a drug is killing the wrong cells or why resistance is forming in real time… it’s not ready to cure cancer. It’s ready to waste another billion dollars. We’re not competing with AI. We’re giving it eyes. Goodbye.

AI is not accelerating nearly fast enough to cure cancer. That is the problem. And it literally cannot without actually seeing what’s happening inside the cell. There’s this new project claiming their models will “solve” cancer by predicting drug responses at scale. Cool story. Except their entire approach still treats the cell like a black box. They’re optimizing for binding affinity while having zero idea what the drug actually does once it’s inside a real, living cell. Meanwhile we’re over here building the tools to watch the cell in real time. No staining or sequencing. Just direct observation of mechanism. You can simulate all you want. But if you can’t see the actual cellular response, you’re just making very confident guesses about something you’ve never actually looked at. We’ve spent two years proving this. The data is there. The difference is night and day. AI is a tool. It’s not the microscope. If your model can’t tell you why a drug is killing the wrong cells or why resistance is forming in real time… it’s not ready to cure cancer. It’s ready to waste another billion dollars. We’re not competing with AI. We’re giving it eyes. Goodbye.

Parmita Mishra

48,203 views • 1 month ago

We are launching A first attempt at building a digital theme park for the Zama Protocol. Our thesis is simple: to accelerate the adoption of FHE, we need cool apps designed for real people. Users don't adopt protocols, they adopt experiences. In most ecosystems, there are very few places that turn participation into status and fun. Usage is transactional, not emotional. Adoption stalls. zashapon is always open, accessible, social by default, and adding infinite attractions over time. It's a playful layer that makes the ecosystem feel like… an ecosystem. The mechanics are simple: - You make FHE transactions by using confidential apps - You get tickets - You can spend them in Zashapon Participation → access → rewards. Zashapon is creating a flywheel: More FHEVM usage → more tickets → more play → more rewards → more users → more builders → more FHEVM usage. Why “gashapon” machines? Because "chance + collection + ritual" is one of the oldest, most reliable engagement loops in games and theme parks. It creates stories users want to repeat. Zashapon uses FHERand to make sure all draws are programmatically fair and confidential. Fun without trust tradeoffs. What you get today: NFTs. What you get tomorrow: anything that can be digitally rewarded — access, perks, whitelists, lore items, partner drops, even real-world benefits. Zashapon is the entertainment / loyalty / identity layer for confidential applications building on top of the Zama Protocol. In physical theme parks: you ride → you collect → you show → you come back with friends. In Zashapon: you use confidential apps → you earn tickets → you draw → you flex / trade / build → you return. We're launching with one Gashapon machine, rewarding the users of Soon there will be many more — one for every new app, partner, and experience built on the Zama Protocol. Digital parks win by expanding attractions continuously. Zashapon is live on Zama testnet right now. Go try it now!

We are launching A first attempt at building a digital theme park for the Zama Protocol. Our thesis is simple: to accelerate the adoption of FHE, we need cool apps designed for real people. Users don't adopt protocols, they adopt experiences. In most ecosystems, there are very few places that turn participation into status and fun. Usage is transactional, not emotional. Adoption stalls. zashapon is always open, accessible, social by default, and adding infinite attractions over time. It's a playful layer that makes the ecosystem feel like… an ecosystem. The mechanics are simple: - You make FHE transactions by using confidential apps - You get tickets - You can spend them in Zashapon Participation → access → rewards. Zashapon is creating a flywheel: More FHEVM usage → more tickets → more play → more rewards → more users → more builders → more FHEVM usage. Why “gashapon” machines? Because "chance + collection + ritual" is one of the oldest, most reliable engagement loops in games and theme parks. It creates stories users want to repeat. Zashapon uses FHERand to make sure all draws are programmatically fair and confidential. Fun without trust tradeoffs. What you get today: NFTs. What you get tomorrow: anything that can be digitally rewarded — access, perks, whitelists, lore items, partner drops, even real-world benefits. Zashapon is the entertainment / loyalty / identity layer for confidential applications building on top of the Zama Protocol. In physical theme parks: you ride → you collect → you show → you come back with friends. In Zashapon: you use confidential apps → you earn tickets → you draw → you flex / trade / build → you return. We're launching with one Gashapon machine, rewarding the users of Soon there will be many more — one for every new app, partner, and experience built on the Zama Protocol. Digital parks win by expanding attractions continuously. Zashapon is live on Zama testnet right now. Go try it now!

Soliton

84,324 views • 7 months ago

Fable 5 comes back！It can now build playable game prototypes. I think it is actually a signal for where AI coding is going. Making a game is not just “write some code.” Even a small browser game needs: game loop；character movement；collision logic；scoring system；UI states；physics tuning；visual feedback；bug fixing；playtesting This is why game prototyping is a great test for AI models. A model cannot fake it with a pretty answer. Either the game runs, or it does not. What impressed me about Fable 5 is that it is useful for the messy middle: turning an idea into mechanics, turning mechanics into code, debugging broken interactions, and iterating until the prototype feels playable. But here is the practical part: I would not use the strongest model for every step. For game building, I would split the workflow: 1. Fable 5 for game design + architecture 2. a fast coding model for routine implementation 3. a vision-capable model for screenshot/UI feedback 4. a cheaper model for docs, test cases, and small fixes 5. fallback when latency, cost, or output quality becomes a problem That is the real AI coding stack. Not “one magic model does everything.” More like: the right model, for the right task, at the right cost, with fallback when things break. This is why I’ve been looking at ZenMux ZenMux. ZenMux gives developers one gateway to access multiple leading AI models, with OpenAI / Anthropic / Google Vertex compatible APIs, cost tracking, quality benchmarks, auto-routing, and compensation when output quality, latency, or throughput falls short. If AI can now make games, the next question is not just “which model is strongest?” It is:how do we manage the whole model workflow Fable 5 shows the creative ceiling. ZenMux is closer to the infrastructure layer you need when AI coding becomes a real production habit.

Fable 5 comes back！It can now build playable game prototypes. I think it is actually a signal for where AI coding is going. Making a game is not just “write some code.” Even a small browser game needs: game loop；character movement；collision logic；scoring system；UI states；physics tuning；visual feedback；bug fixing；playtesting This is why game prototyping is a great test for AI models. A model cannot fake it with a pretty answer. Either the game runs, or it does not. What impressed me about Fable 5 is that it is useful for the messy middle: turning an idea into mechanics, turning mechanics into code, debugging broken interactions, and iterating until the prototype feels playable. But here is the practical part: I would not use the strongest model for every step. For game building, I would split the workflow: 1. Fable 5 for game design + architecture 2. a fast coding model for routine implementation 3. a vision-capable model for screenshot/UI feedback 4. a cheaper model for docs, test cases, and small fixes 5. fallback when latency, cost, or output quality becomes a problem That is the real AI coding stack. Not “one magic model does everything.” More like: the right model, for the right task, at the right cost, with fallback when things break. This is why I’ve been looking at ZenMux ZenMux. ZenMux gives developers one gateway to access multiple leading AI models, with OpenAI / Anthropic / Google Vertex compatible APIs, cost tracking, quality benchmarks, auto-routing, and compensation when output quality, latency, or throughput falls short. If AI can now make games, the next question is not just “which model is strongest?” It is:how do we manage the whole model workflow Fable 5 shows the creative ceiling. ZenMux is closer to the infrastructure layer you need when AI coding becomes a real production habit.

Rachel🥥

60,942 views • 23 days ago

This guy cracked the code on AI-powered fashion ecommerce using synthetic face technology and now pulls $50,000 to $150,000 per month from two Shopify stores without paying a single real model. He got tired of watching DTC fashion brands burn $20,000 monthly on photoshoots while their competitors tested 40 product angles in the same timeframe, so he built a system that generates hyperrealistic fashion content using his gaming PC and real-time AI masks instead of studios, contracts, or casting calls. His monthly profit hit $150,000 last month from just 2 stores and organic TikTok traffic, while traditional fashion brands cap out at $30K after paying models $400 to $800 per shoot and studio rentals of $200 to $500 per session. Here is the exact breakdown: → Real-time synthetic face technology becomes the only tool you need, but most people butcher the setup by skipping motion sync calibration in the first 30 seconds → Product selection comes first, and if you mess this up nothing saves it. Stick to women's accessories (bags, sunglasses, jewelry) because that is where organic TikTok engagement lives → Avatar casting is not random. You build one consistent AI face that repeats across all content so your audience recognizes the "model" and trusts the brand continuity → You are picking who your customer projects onto, not who looks expensive. That is your positioning baked into the face → Motion capture runs before generation, and this is what kills the uncanny valley effect that destroys watch time in 4 seconds → You mirror your own gestures through webcam: wave, chin tap, finger point, shoulder dance. The AI mask tracks every micro-movement and applies it to the generated face in real time → Batching is the move 94 percent skip: same outfit base, multiple product swaps, one recording session. No re-shooting, no model schedules, no usage rights negotiations → The system generates 3 to 5 TikToks before lunch, while traditional brands test 2 per week and wonder why their conversion rates are stuck at 0.8 percent The economics are stupid: each video costs him $0 in talent fees, pulls 1.5 million views organically, converts at 0.03 percent into 450 orders at $45 to $60 retail with $30 to $45 margin per sale. That is $15,750 profit per viral video, while fashion brands pay $1,200 per shoot and net $3,000 after ads. The key move nobody talks about: you cannot skip the motion synchronization test. If you generate the AI face without mirroring your own natural gestures first, the avatar moves like a mannequin. The blinks lag. The smile timing breaks. The whole thing screams "synthetic face technology" and your hook rate dies at 1.1 seconds. His system records him doing the exact dance trend first, so the AI mask inherits human timing, natural head tilts, and spontaneous energy that reads as a real creator showing off a product find, not a rendered advertisement. One accessories store generated 10 variants of the same handbag reveal in 18 minutes with different outfits, different backgrounds, different trend audios, and found the winner in 72 hours without spending $6,000 on influencer gifting. They were previously paying $800 per UGC creator and burning $4,800 per week on content that plateaued at 40K views. Now they spend $0 for 10 variants and their cost per acquisition dropped from $62 to $18. UGC agencies now panic because their entire margin was built on talent scarcity, and this removes the human bottleneck. The outfit changes between clips like a wardrobe filter. The lighting matches bedroom setups. The hand gestures sync with beat drops. No casting call. No model release. No location permits. Just a webcamera, a real-time AI mask, and the discipline to batch-test product angles before you commit ad spend to one creative.

Shade

20,010 views • 2 months ago

$Introducing Glidepath. A new way for builders on Bankr to take profit -- without nuking their own chart, or their reputation. The problem: Builders earn fees in their own token. The second they sell into the pool, the chart craters, holders get wrecked, and trust evaporates. And they torch their own long-term upside doing it. First -- what Glidepath is not: It doesn't pull liquidity. It never touches your pool's LP. Pulling liquidity makes trading your token inefficient and unappealing. It's your own tokens, fed back into the pool in slices so small the market barely registers them, each one sized by the Bankr AI agent to live conditions. Why that's healthy for the chart, not harmful: Every slice is a tiny fraction of pool depth, spread over time. Organic buy volume absorbs it, price can keep trending instead of taking a wick. A small, steady, absorbable flow is nothing like a full clip. It actually gets better. Once "the dev might dump" is off the table, buyers price in less risk. The overhang that caps every launch disappears. Less rug risk → stronger bid. Committing to a Glidepath can be bullish. And it's not opt‑in. Selling your fee token straight into the pool through Bankr is now turned off -- Glidepath is the only way to sell it on Bankr. So "the dev might dump" stops being a promise holders have to trust, and becomes a rule they can see. Credible commitment -- enforced, not just offered. And here's the part builders sleep on: Before you commit, Glidepath shows what that same stack is worth at higher market caps. You don't have to dump to fund your project. Grind the coin up, and the same tokens fund you many times over. Your treasury grows with your chart, not against it. Once you commit: → tokens are locked to a vesting wallet → after a short heads-up window (48hr), they exit in small slices using the AI generated sell plan → each slice capped to a fraction of real liquidity -- the AI can size under the cap, never over And it's all in the open. Your token page shows a live exit plan for everyone to see -- committed, sold, remaining -- with the exact timing fuzzed so it can't be front-run. Holders see a capped, transparent glide. No hidden float. No 3am chart nuke. Bottom line: Creators -- take profit on your terms, chart and reputation intact. Holders -- "the dev might dump" becomes a known, capped, visible number known up front. For once, you and your holders want the exact same thing: number go up. This is what launching on Bankr should mean: credible commitment, built in. Glidepath now live in your Bankr terminal$

Introducing Glidepath. A new way for builders on Bankr to take profit -- without nuking their own chart, or their reputation. The problem: Builders earn fees in their own token. The second they sell into the pool, the chart craters, holders get wrecked, and trust evaporates. And they torch their own long-term upside doing it. First -- what Glidepath is not: It doesn't pull liquidity. It never touches your pool's LP. Pulling liquidity makes trading your token inefficient and unappealing. It's your own tokens, fed back into the pool in slices so small the market barely registers them, each one sized by the Bankr AI agent to live conditions. Why that's healthy for the chart, not harmful: Every slice is a tiny fraction of pool depth, spread over time. Organic buy volume absorbs it, price can keep trending instead of taking a wick. A small, steady, absorbable flow is nothing like a full clip. It actually gets better. Once "the dev might dump" is off the table, buyers price in less risk. The overhang that caps every launch disappears. Less rug risk → stronger bid. Committing to a Glidepath can be bullish. And it's not opt‑in. Selling your fee token straight into the pool through Bankr is now turned off -- Glidepath is the only way to sell it on Bankr. So "the dev might dump" stops being a promise holders have to trust, and becomes a rule they can see. Credible commitment -- enforced, not just offered. And here's the part builders sleep on: Before you commit, Glidepath shows what that same stack is worth at higher market caps. You don't have to dump to fund your project. Grind the coin up, and the same tokens fund you many times over. Your treasury grows with your chart, not against it. Once you commit: → tokens are locked to a vesting wallet → after a short heads-up window (48hr), they exit in small slices using the AI generated sell plan → each slice capped to a fraction of real liquidity -- the AI can size under the cap, never over And it's all in the open. Your token page shows a live exit plan for everyone to see -- committed, sold, remaining -- with the exact timing fuzzed so it can't be front-run. Holders see a capped, transparent glide. No hidden float. No 3am chart nuke. Bottom line: Creators -- take profit on your terms, chart and reputation intact. Holders -- "the dev might dump" becomes a known, capped, visible number known up front. For once, you and your holders want the exact same thing: number go up. This is what launching on Bankr should mean: credible commitment, built in. Glidepath now live in your Bankr terminal

Bankr

97,592 views • 1 month ago

Today marks General Availability of AgentCore, a set of infrastructure building blocks for developers and companies to build secure, scalable agents. When we first started AWS, the vast majority of developers were spending most of their time on the undifferentiated heavy lifting of infrastructure instead of what differentiated their feature. So, we solved that problem by building primitive building blocks like compute and storage and database that would allow teammates and customers to quickly build and deploy new experiences without having to reinvent the wheel each time. We realized the same thing was happening with AI agents. It's too difficult and it's slowing customers down. That's why we created AgentCore, a set of services to build, deploy, and operate highly capable agents using any framework or model, with enterprise-grade security and scalability. These building blocks (like serverless secure runtime, memory, observability, a gateway that does MCP translation, etc) help customers tackle some of the biggest challenges of going from prototype to production, much more quickly, securely, and scalably. AgentCore has been in preview for several weeks, and customers have been quite excited about it. The AgentCore SDK has already been downloaded over a million times and we're seeing transformative results, such as Cohere Health expecting to reduce medical review times by 30-40% in highly regulated healthcare, and teams at Cox Automotive and Experian are embracing its flexibility to deploy and operate agents at scale. Inside Amazon, our Amazon Devices Operations & Supply Chain team is using AgentCore to develop an agentic manufacturing approach where AI agents work together to automate manual processes – turning what used to be days of engineering time into processes that take under an hour with high precision. Just like AWS changed how companies build and scale applications, we believe AgentCore will do the same for AI agents, enabling the next generation of innovation.

Today marks General Availability of AgentCore, a set of infrastructure building blocks for developers and companies to build secure, scalable agents. When we first started AWS, the vast majority of developers were spending most of their time on the undifferentiated heavy lifting of infrastructure instead of what differentiated their feature. So, we solved that problem by building primitive building blocks like compute and storage and database that would allow teammates and customers to quickly build and deploy new experiences without having to reinvent the wheel each time. We realized the same thing was happening with AI agents. It's too difficult and it's slowing customers down. That's why we created AgentCore, a set of services to build, deploy, and operate highly capable agents using any framework or model, with enterprise-grade security and scalability. These building blocks (like serverless secure runtime, memory, observability, a gateway that does MCP translation, etc) help customers tackle some of the biggest challenges of going from prototype to production, much more quickly, securely, and scalably. AgentCore has been in preview for several weeks, and customers have been quite excited about it. The AgentCore SDK has already been downloaded over a million times and we're seeing transformative results, such as Cohere Health expecting to reduce medical review times by 30-40% in highly regulated healthcare, and teams at Cox Automotive and Experian are embracing its flexibility to deploy and operate agents at scale. Inside Amazon, our Amazon Devices Operations & Supply Chain team is using AgentCore to develop an agentic manufacturing approach where AI agents work together to automate manual processes – turning what used to be days of engineering time into processes that take under an hour with high precision. Just like AWS changed how companies build and scale applications, we believe AgentCore will do the same for AI agents, enabling the next generation of innovation.

Andy Jassy

24,990 views • 9 months ago

A CHINESE GUY STOPPED PAYING WEB DESIGNERS $1,800 PER LANDING PAGE AND BUILT THE SAME KIND OF SITE WITH CLAUDE CODE IN ONE AFTERNOON FOR LESS THAN $70 claude was not just writing code. it was doing the job of a designer, copywriter and frontend dev in one window. in 4 hours, one rough idea turned into the layout, color system, font choices, page sections and the full html/css/js build the real trick was the references. he gave claude 5 screenshots, made it ask 7 questions first, then pushed it toward one clear visual direction instead of accepting another generic ai-looking template the first version already looked solid. the second pass made it feel expensive: better typography, a darker palette, mobile cleanup, cursor effects, 6 micro-interactions and custom hero visuals his old workflow was burning $1,200 on design, $500 on frontend work and another $150 on small fixes every time he needed a new page. now the whole test costs less than $70 and the site still looks like something a $5,000 agency would ship the edge is not “ai builds websites.” the edge is that one person can now brief, critique, polish and launch in one afternoon without waiting 10 days for a designer to send version one

A CHINESE GUY STOPPED PAYING WEB DESIGNERS $1,800 PER LANDING PAGE AND BUILT THE SAME KIND OF SITE WITH CLAUDE CODE IN ONE AFTERNOON FOR LESS THAN $70 claude was not just writing code. it was doing the job of a designer, copywriter and frontend dev in one window. in 4 hours, one rough idea turned into the layout, color system, font choices, page sections and the full html/css/js build the real trick was the references. he gave claude 5 screenshots, made it ask 7 questions first, then pushed it toward one clear visual direction instead of accepting another generic ai-looking template the first version already looked solid. the second pass made it feel expensive: better typography, a darker palette, mobile cleanup, cursor effects, 6 micro-interactions and custom hero visuals his old workflow was burning $1,200 on design, $500 on frontend work and another $150 on small fixes every time he needed a new page. now the whole test costs less than $70 and the site still looks like something a $5,000 agency would ship the edge is not “ai builds websites.” the edge is that one person can now brief, critique, polish and launch in one afternoon without waiting 10 days for a designer to send version one

Gipp 🦅

295,173 views • 1 month ago