{"ok":true,"generated_at":"2026-09-19T07:08:57.923Z","influencers":[{"name":"David Shapiro","url":"https://www.youtube.com/results?search_query=David+Shapiro+AI"},{"name":"David Ondrej","url":"https://www.youtube.com/results?search_query=David+Ondrej+AI"},{"name":"AI Grid","url":"https://www.youtube.com/results?search_query=AI+Grid"},{"name":"Matt Wolfe","url":"https://www.youtube.com/results?search_query=Matt+Wolfe+AI"},{"name":"Matthew Berman","url":"https://www.youtube.com/results?search_query=Matthew+Berman+AI"},{"name":"Wes Roth","url":"https://www.youtube.com/results?search_query=Wes+Roth+AI"},{"name":"Dan Martell","url":"https://www.youtube.com/results?search_query=Dan+Martell+AI"},{"name":"Jeff Su","url":"https://www.youtube.com/results?search_query=Jeff+Su+AI"},{"name":"Futurepedia","url":"https://www.youtube.com/results?search_query=Futurepedia+AI"},{"name":"Nate B. Jones","url":"https://www.youtube.com/results?search_query=Nate+B+Jones+AI"},{"name":"Network Chuck","url":"https://www.youtube.com/results?search_query=Network+Chuck+AI"},{"name":"Marina Wyss","url":"https://www.youtube.com/results?search_query=Marina+Wyss+AI"}],"items":[{"videoId":"roHLWMmDBsc","title":"How to Build AI Agent Skills That Actually Work (FREE)","description":"Host Your AI Agent in Hostinger 👉 https://www.hostg.xyz/SHJvn In this video, I show how to create a custom SOUL.md and SKILL.md, deploy Hermes on a Hostinger server, connect it through Telegram, build a repeatable research skill, test and refine it, and schedule it to deliver consistent work automatically while you sleep. Grab The Hermes Agent Playbook 👉 https://parker-prompts.com/aiagentskills Create Your Own Agent 👉 https://www.ordain.host/ Go from ABSOLUTE ZERO to using AI like the people who do this for a living👇 https://www.skool.com/ai-fluency I'm Parker. I started this YouTube Channel with the goal to learn more about AI myself and to then pass on the knowledge to anyone willing to listen. let's work together: partnerships @ parker-prompts.com","channelTitle":"Parker Prompts","channelId":"UCaNk22cLid93kifuVbVapcQ","channelSubscriberCount":131000,"channelThumbnail":"https://yt3.ggpht.com/rmKIUlPL24nBNHrCDwupmmA5ilpfMDRlm8mP-bnbIvantod3Tzxo-ezrZjAiMoOW4iVezlEUIA=s88-c-k-c0x00ffffff-no-rj","viewCount":8041,"likeCount":0,"durationMinutes":12.45,"published_at":"2026-09-18T15:04:25Z","ageHours":16.075811944444446,"thumbnail":"https://i.ytimg.com/vi/roHLWMmDBsc/hqdefault.jpg","url":"https://www.youtube.com/watch?v=roHLWMmDBsc","rankScore":3.4273},{"videoId":"wTl4HvEiY5g","title":"Building Your First AI Agent Is Actually Pretty Easy","description":"How To Build An AI Agent With Claude Code (Tutorial) Welcome to Dawnspire Academy everyone! If you, like Chip, want to setup your first AI Agent or you just want to make it actually useful - you're in the right place! Claude Code is what we're using as the agent - the setup is really simple and straight to the point... or is it? By the end of this claude code tutorial you'll know exactly how to use it and all the cheeky little tricks I used to set it up properly. MD file: https://drive.google.com/file/d/1AcSGJOaN14oJl-WjhbJ7PP6bAu_xoXd8/view?usp=sharing","channelTitle":"Theo - Dawnspire Academy","channelId":"UCIBN_ixxnop1MGbeGVvNUSA","channelSubscriberCount":13400,"channelThumbnail":"https://yt3.ggpht.com/xyineyqF6IgKe9IG-F0s-n1zMtC4QkwBcSWjlSRdLt2zbUenOrGT2BXHIPZUa5e6zxeqApF_L2k=s88-c-k-c0x00ffffff-no-rj","viewCount":5599,"likeCount":0,"durationMinutes":9.433333333333334,"published_at":"2026-09-18T11:50:37Z","ageHours":19.305811944444443,"thumbnail":"https://i.ytimg.com/vi/wTl4HvEiY5g/hqdefault.jpg","url":"https://www.youtube.com/watch?v=wTl4HvEiY5g","rankScore":3.0933},{"videoId":"VFVGr2edzrc","title":"Never Give Your AI a Shell #aisecurity #homelab #selfhosted #aiagents #devops #sysadmin #shorts","description":"If you connect an AI assistant to your servers, this is the boundary you should not skip. HERA never receives a shell. It gets narrow, predefined capabilities backed by trusted code. Even if the model asks for something dangerous, and even if I ask through chat, the read-only connector cannot change the system. For future actions, the rule stays simple: the model proposes, a human approves, and deterministic code executes. Explore HERA: https://github.com/theeromi/hera More projects: https://codewithromi.com Instagram: @codewithromi #aisecurity #homelab #selfhosted #aiagents #devops #sysadmin #shorts","channelTitle":"CodeWithRomi","channelId":"UC7NHfdcSM28NsfaQ4uAn4oA","channelSubscriberCount":4980,"channelThumbnail":"https://yt3.ggpht.com/U_fAsEGB6Rhuh2GgIm6vMtlS2ywIpk0whsM-MBqu1wYTCdN0eIxgJRPR5ppNtA2EJhp45CeHmw=s88-c-k-c0x00ffffff-no-rj","viewCount":933,"likeCount":1,"durationMinutes":0.6666666666666666,"published_at":"2026-09-18T20:00:16Z","ageHours":11.144978611111112,"thumbnail":"https://i.ytimg.com/vi/VFVGr2edzrc/hqdefault.jpg","url":"https://www.youtube.com/watch?v=VFVGr2edzrc","rankScore":2.6182},{"videoId":"rjNLeB_q7DM","title":"6 agentes de IA crean una campaña desde una foto","description":"Comenta “EQUIPO” y te envío el vídeo completo donde construyo este sistema desde cero. He creado seis agentes de IA y cada uno tiene un único trabajo: — Director — Analista — Guionista — Arte — Cámara — Redactor Yo solamente les entrego una fotografía del producto. Cuando vuelvo, tengo preparada la ficha del producto, el ángulo de venta, el guion plano por plano, un personaje consistente, un carrusel de imágenes y dos vídeos en castellano. La diferencia está en no pedirle a un único agente que lo haga absolutamente todo. El director reparte el trabajo y cada especialista ejecuta únicamente su parte. Montarlo lleva una tarde. Después puedes reutilizar el sistema con cualquier producto. Comenta EQUIPO y te envío el vídeo de YouTube donde construyo la fábrica completa paso a paso. #agentesia #automatizacion #inteligenciaartificial #marketingdigital #creaciondecontenido","channelTitle":"Pau Rodríguez | Inteligencia Artificial","channelId":"UCHsO9q47lb3mp0lgzdtvySw","channelSubscriberCount":3990,"channelThumbnail":"https://yt3.ggpht.com/xsONAFC4pLuu51dsU_xpCabkXr_DzqSNyXDc18n1J4WjqjcIkCGHVMs78YjGoNSxt-zol708_w=s88-c-k-c0x00ffffff-no-rj","viewCount":307,"likeCount":18,"durationMinutes":1.35,"published_at":"2026-09-18T17:00:03Z","ageHours":14.148589722222223,"thumbnail":"https://i.ytimg.com/vi/rjNLeB_q7DM/hqdefault.jpg","url":"https://www.youtube.com/watch?v=rjNLeB_q7DM","rankScore":2.5247},{"videoId":"ziTgfxqbvY0","title":"Meta AI Agent in WhatsApp Business App | Setup Guide","description":"Learn how to set up the Meta AI agent in the WhatsApp Business app to answer customer questions and support your business around the clock. This quick walkthrough covers: • Adding your business details • Connecting your product catalogue • Adding payment, shipping, and return information • Setting up customer details and appointments • Customising response settings, knowledge, and instructions • Testing your AI agent before using it Subscribe to GreenAds Global for more WhatsApp Business app tips and Meta AI updates. Learn More:https://www.greenadsglobal.com/post/meta-business-agent-for-whatsapp-business-app?utm_source=youtubeshorts&utm_medium=metabusinessaiagentmeeting&utm_campaign=metabusinessaiagentforwhatsappbusiness Book a Meeting:https://zcal.co/greenadsglobal?utm=&utm_source=youtubeshorts&utm_medium=metabusinessaiagentmeeting&utm_campaign=metabusinessaiagentforwhatsappbusiness Feature availability and menu names may vary by account, region, and app version. #WhatsAppBusiness #MetaAI #Shorts","channelTitle":"GreenAds Global","channelId":"UCAaOdsnaWilyCuac0cUyA0A","channelSubscriberCount":358,"channelThumbnail":"https://yt3.ggpht.com/sl0QmpztHQ7PNXQC_TO0CiAKfXOmJ0-KfbcTs791InVq5pJi9yARctG2wosDZa0H0ddFQ1cB=s88-c-k-c0x00ffffff-no-rj","viewCount":331,"likeCount":6,"durationMinutes":1.0833333333333333,"published_at":"2026-09-18T14:15:19Z","ageHours":16.894145277777778,"thumbnail":"https://i.ytimg.com/vi/ziTgfxqbvY0/hqdefault.jpg","url":"https://www.youtube.com/watch?v=ziTgfxqbvY0","rankScore":2.1944},{"videoId":"-H3ts5yVHQU","title":"Meta Muse vs OpenClaw vs Hermes: 3 AI Agents Compared","description":"Meta Muse vs OpenClaw vs Hermes : 3 AI agents, 3 very different approaches. We break down the key differences in environment, memory, tools, models, automation, control, and customization. Muse focuses on managed personal AI. OpenClaw focuses on infrastructure and control. Hermes focuses on extensibility, tools, memory, and automation. This is the quick version, watch the full detailed comparison on Kernel Cast for the complete breakdown. #aiagents #muse #openclaw #hermes #metaai #artificialintelligence #aiautomation #ai #kernelcast","channelTitle":"Kernel Cast","channelId":"UC3PhOJM5EpgVJgurVPrnqiw","channelSubscriberCount":1000,"channelThumbnail":"https://yt3.ggpht.com/uNq040yH2RolRJWi1742AbRSusymx80H2JyQyCZz3gLW0uUkSlLByxiNKjYHEih3AFqkYhreig=s88-c-k-c0x00ffffff-no-rj","viewCount":171,"likeCount":2,"durationMinutes":1.8333333333333335,"published_at":"2026-09-18T18:00:05Z","ageHours":13.148034166666667,"thumbnail":"https://i.ytimg.com/vi/-H3ts5yVHQU/hqdefault.jpg","url":"https://www.youtube.com/watch?v=-H3ts5yVHQU","rankScore":2.0751},{"videoId":"njPCrr-QQ48","title":"AI Agents Are Changing How We Work 🚀","description":"📞Contact: +918130537300 🌍Website: www.skillsforeveryone.com Follow us to engage and know more: Instagram: https://www.instagram.com/skills4everyone_official/ WhatsApp for Admission: ‪https://wa.me/919056777027 Youtube: www.youtube.com/@SkillsforEveryone_com Our mission is to prepare you for the future: At the heart of our school is our unwavering mission to prepare you for the future. We understand that the landscape of education is evolving rapidly, and our programs are designed to equip you with the skills and knowledge needed to thrive in tomorrow's world. #AIAgents #ChatGPT #AI #GenerativeAI #Automation #FutureOfAI #Cloud #Networking #IT #Technology","channelTitle":"Skills For Everyone","channelId":"UC_wn14LrWAscLnmfSYB0qyw","channelSubscriberCount":4130,"channelThumbnail":"https://yt3.ggpht.com/ZBofZLoVmehpn9B3oJydtcuO-nmsMArh9uDF4mSKhCvZpF-GCFSTJO3xWu8WxRIWvcrPNMm-uw=s88-c-k-c0x00ffffff-no-rj","viewCount":77,"likeCount":3,"durationMinutes":1.65,"published_at":"2026-09-19T05:18:46Z","ageHours":1.8366452777777778,"thumbnail":"https://i.ytimg.com/vi/njPCrr-QQ48/hqdefault.jpg","url":"https://www.youtube.com/watch?v=njPCrr-QQ48","rankScore":2.0651},{"videoId":"W5m3z4QnyN0","title":"🤯 Your Agentic RAG Is Useless Without These 2 Models!","description":"Building an Agentic RAG system? 🤖📚 You need to load the right LLM and embedding model before your agent can understand queries and retrieve relevant information. In this Short, see how LLMs + embedding models fit into an Agentic RAG pipeline and why both are essential for building effective AI applications. 🚀 Build Your First RAG App with LlamaIndex — Hands-On Project: https://youtu.be/Q7TmIlZdWf0 In the full tutorial, you'll build a real RAG application with LlamaIndex, including document loading, chunking, embeddings, retrieval, and LLM-powered responses. Subscribe for more AI Engineering, Agentic RAG, LlamaIndex, LangChain & AI Agents! 🚀 #AgenticRAG #RAG #LlamaIndex #AIEngineering #LLM #Embeddings #GenerativeAI #AIAgents #Python #Shorts","channelTitle":"Tech With Mala","channelId":"UC2_X9JyHWEPtSatHfVmglcA","channelSubscriberCount":3680,"channelThumbnail":"https://yt3.ggpht.com/hh5BfUZc53W1SDW5BQMfqWgK574PFCvQUy_Bqy5KzLQriAG64yyV3PMoAEnahBRQ_meYJfrSdRU=s88-c-k-c0x00ffffff-no-rj","viewCount":63,"likeCount":2,"durationMinutes":0.48333333333333334,"published_at":"2026-09-18T23:00:34Z","ageHours":8.139978611111111,"thumbnail":"https://i.ytimg.com/vi/W5m3z4QnyN0/hqdefault.jpg","url":"https://www.youtube.com/watch?v=W5m3z4QnyN0","rankScore":1.9803},{"videoId":"DdoB46AhFUg","title":"The Internet Is Changing. AI Agents Will Do the Work.","description":"Today, you use the internet. Tomorrow, AI might use it for you. Instead of searching, comparing, and clicking through websites yourself, AI agents can handle the task for you. Tell it what you want. Let it do the work. What would you ask an AI agent to do? 👀 Comment “AGENT” if you want to explore it. 📲 Follow AutoFlow on Instagram: @autoflow.work #AIAgents #ArtificialIntelligence #AI #FutureOfAI #Automation #AITools #Technology #AutoFlow #Shorts","channelTitle":"AutoFlow","channelId":"UCRTuHjGYb1yyIFZ311WFoQQ","channelSubscriberCount":5,"channelThumbnail":"https://yt3.ggpht.com/9nMrJi9mCBbouCXC2kvCnlYsSFG-DdA1mk23rr9Wc9e0vTFiIkGlkzT2dqyTkPyDOo8NKuUh=s88-c-k-c0x00ffffff-no-rj","viewCount":406,"likeCount":8,"durationMinutes":0.5166666666666667,"published_at":"2026-09-18T16:25:53Z","ageHours":14.718034166666667,"thumbnail":"https://i.ytimg.com/vi/DdoB46AhFUg/hqdefault.jpg","url":"https://www.youtube.com/watch?v=DdoB46AhFUg","rankScore":1.8207},{"videoId":"cvYdYJxwdXU","title":"A 9B model on one GPU built this entire site  #ai #aiagents #localai #aivideo #github #automation","description":"This entire React website was built by a 9-billion-parameter model running on one 12GB graphics card. No cloud. No API. It also fixed a real bug by changing a single line of code. The model is OxCoder-9B — built by OrionLLM on Qwen3.5-9B, Apache 2.0, 6GB file. MEASURED ON THIS MACHINE (RTX 3060 12GB) • 37 tok/s steady decode • 7,899 MiB peak VRAM • 0 console errors at 1440px WHAT IT DID • Task 1 — LRU cache: correct algorithm (doubly-linked list + hash map), then crashed because it named a class `node` and used it as `Node`. One wrong capital letter. • Task 2 — real bug fix: found an off-by-one pagination bug and changed ONE line. Tests 2/5 → 5/5. • Task 3 — this site: 13 React files, dark theme, 3D hero, six sections. It needed 8 repair rounds, and twice the build PASSED while the page was blank because it deleted components to compile. The full breakdown — all three tasks, the failures, and the honest verdict — is on my channel. #localai #llm #coding #ai #rtx3060 #OxCoder #qwen #TheLogicRoom #shorts **Q: Can a 9B model build a website on a 12GB GPU?** Yes, with qualification. In this test OxCoder-9B produced 13 React files — a dark landing page with a navigation bar, a Three.js hero, six project cards, a team section, services, a contact form and a footer — that built and ran with zero console errors at 1440px. It required eight build-fix cycles driven by real compiler output, and at two points it deleted component implementations to make the build pass. **Q: What is OxCoder-9B?** A 9-billion-parameter coding model by OrionLLM, built on Qwen3.5-9B under the Apache 2.0 licence. It targets agentic coding and multi-file edits, has a 262,144-token context window, and ships as a ~6GB GGUF at Q5_K_M. **Q: How fast is a 9B coding model on an RTX 3060?** About 37 tokens per second sustained decode, with peak VRAM around 7,899 MiB during generation. **Q: What goes wrong when a local model writes a React app?** Not the code itself — the cross-file integration. Observed failures: a case-sensitive identifier mismatch that crashed a correct LRU cache; importing a stylesheet that was never created; using default exports in two files while using named exports in eleven; and referencing a variable that was never imported. Each was a small integration error, and each was fixable once the real compiler message was supplied.","channelTitle":"The Logic Room","channelId":"UChlkraELSM_KE8CQ0W4vwGg","channelSubscriberCount":215,"channelThumbnail":"https://yt3.ggpht.com/_3svBV8-AtAk49-nMbhQ8dfWf9t8XNE40VNZ7RPvW2ciiZZrlDmRxwpuLtWKLxf52FtAHt22=s88-c-k-c0x00ffffff-no-rj","viewCount":128,"likeCount":0,"durationMinutes":1,"published_at":"2026-09-18T17:15:50Z","ageHours":13.885534166666666,"thumbnail":"https://i.ytimg.com/vi/cvYdYJxwdXU/hqdefault.jpg","url":"https://www.youtube.com/watch?v=cvYdYJxwdXU","rankScore":1.7444}],"fallback":false}