Poo AI Lab @pooaa467

@ai_with_jyoti Book marking

7 AI agents topics to learn

@KeriRetro @Divyyanshishrma Wrong

How to write a Good Prompt: *1: Role* 🙇‍♂️ — who you are and who you're doing this for _(I run an AI automation agency where we help business owners implement automations into their systems)_ *2: Goal* 🎯 — what you want Claude/GPT to actually do - write or build _(build a dashboard that I can connect to n8n, VAPI, and Relevance that shows all the usage of API costs and gives me the actual amount getting used every month)_ *3: Context* 📄 — anything to help Claude/GPT understand your situation better that it would otherwise have to guess _(I have all the API keys, I am not a developer, and I have 5 clients)_ *4: Format* 📂 — what the output should look like _(the dashboard should be clean, minimalistic, just black and white)

Where AI can help a business (and u can charge for it) 1: Marketing 2: Sales 3: Product 4: Customer Support/Success 5: IT 6: Recruitment 7: HR 8: Finance 9: Legal

2nd brain which one is better 1. Notion 2. Obsidian 3. Google drive

🚀 One GitHub repo every developer should know: Public APIs — a massive collection of free APIs you can use to build, test, and experiment with real projects. You’ll find APIs across: • Business • Finance • Weather • Games • Sports • AI • News • Development & more Perfect for beginners building their first API project or AI engineers looking for data sources. 🔗 github.com/public-apis/pu… Save this. Bookmark it. Build something. 🔥 #AIEngineering #APIs #GitHub #Programming #Developers

AI models have an expiry date. ⚠️ If you're building with LLMs in 2026, don't treat a model like a permanent software dependency. Models can be deprecated because of: • Infrastructure & GPU costs • Better architectures • Security & compliance • Economics And there's another hidden cost: tokens. Input tokens are generally cheaper than output tokens, so inefficient prompts, long contexts, and repeated agent loops can quietly burn your API budget. The real AI engineering skill isn't just knowing how to call an LLM. It's knowing how to build systems that survive model changes, pricing changes, and deprecations. Build for change. Not for one model. 🚀 #AIEngineering #LLM #GenerativeAI #AI #AIAgents

AI watermark removal just got open-sourced. This GitHub project can strip AI provenance marks from text and files, including: • Invisible Unicode & text watermarks • C2PA / EXIF / XMP metadata • PNG, JPEG, SVG, PDF, DOCX, HTML & Markdown • Optional AI-based text rewriting A fascinating project to study if you're building AI agents, content tools, or AI provenance systems. github.com/guillaumemeyer…

Ronin @DeRonin_

3 days ago

to make the easiest $100k MRR now, start building anti-AI detection systems and watermark removers demand is proved after this release.

39 7 212 27K 124

to make the easiest $100k MRR now, start building anti-AI detection systems and watermark removers demand is proved after this release.

NIK @ns123abc

a week ago

🚨 JUST IN: Claude models will now have invisible watermarks embedded in ALL text, and ALL metadata attached to files…

keyboard_arrow_left Previous keyboard_arrow_right Next

1K 3K 30K 13.7M 14K

@etanddd @thestatusai Oh man , opening new insight for beginner ugc creator connected

@tankots Being an ai company u hired a human why can't build ai agent to scrap the viral content

Andrew Ng’s AI Engineering Skills Map is a great reality check. The 4 skills every developer should build: 1. Build & deploy AI applications 2. Master software engineering fundamentals 3. Use coding agents effectively 4. Shape what gets built The future isn’t just about coding faster. It’s about knowing what to build, how to guide AI to build it, and how to verify it actually works. Continuous learning is the skill underneath all four.

Andrew Ng @AndrewYNg

4 days ago

New: A map of the most important skills in AI Engineering. x.com/i/article/2088…

374 4K 22K 5.5M 48K

Stop trying to build a SaaS as your first project. Build a tool for a problem you personally face. The better path: Problem you face → Personal tool → Team uses it → Improve it → More users → Product → SaaS Don't start with monetization. Start by building something people actually want to use. Your first app should teach you to solve problems, not sell subscriptions.

GLM-5.3 just dropped. Z.ai says it’s a major leap for open-weights coding agents: • 50% better than GLM-5.2 on Z.ai Code Bench • 28.3 vs 4.6 on Terminal-Bench 3.0 • 84.5% on CyberGym • ExploitBench: 24.4% → 54.4% • 2,436 vulnerabilities found across 269 projects • Open weights coming in ~2 weeks The interesting part? The model didn’t just get better at coding. Its cyber capabilities emerged as post-training scaled. AI agents are moving from “write code” to “do the entire job.”

Z.ai @Zai_org

4 days ago

Introducing GLM-5.3: Built to Code. Ready for Cyber Defense. - Top-tier coding and agentic capabilities, achieved through post-training on the 743B base model - A major leap in cybersecurity, setting a new standard among open models Tech Blog: z.ai/blog/glm-5.3

933 2K 19K 5.5M 3K

Introducing GLM-5.3: Built to Code. Ready for Cyber Defense. - Top-tier coding and agentic capabilities, achieved through post-training on the 743B base model - A major leap in cybersecurity, setting a new standard among open models Tech Blog: z.ai/blog/glm-5.3

@iamfakhrealam Seriously its usefull to participate in. Hackathon to create great mvp

Top 10 AI claude side hustle Monetize a Skill You Already Have LinkedIn & Email Ghostwriting Outbound Lead Generation AI Employees & Automations Business Data & CRM Automation Website Building MVP Development Faceless AI Content Building Claude Skills Selling Prompt Packs & Templates

@Dozie101 @daveylan2 Well said learn the game , don't blame