Lu Han

@Davidstrolder Just give it a try. Really good.

Stop reading books and start talking to them 🗣️ • Get life advice • Debate their deepest ideas • Vent about your problems

@heyblake jobtact.app

Drop your project URL Let’s drive some traffic

@zarazhangrui Everybody say start with pain points, but you say start with technology. I think your point of view is reasonable.

Am I the only one? AI tools' "free trial" is the new dark pattern. - Auto-charge by default - No reminder before renewal - Cancel button hidden 3 clicks deep Lost $100+ so far to forgotten subscriptions. Does anyone know a tool to manage this mess?

@JulianGoldieSEO yes

⚡ I BUILT AN AI AGENT THAT RUNS MY SALES FOR ME This thing works 24/7… even when I’m offline. Last night alone: → Followed up with 86 leads → Closed 4 deals automatically No human needed. Here’s what you’ll learn: • How to build an AI voice agent for appointment confirmations • How to connect your calendar to a dialer system • How to enable real-time responses with GPT-5 • How to access the full template (free) • How to set everything up without coding Like + reply “YES” & I’ll send you the FULL workflow + setup FREE!

People who say "AI writes my posts in 5 minutes" are making trash content. A good AI post still takes me 40 minutes. You must write the core idea yourself. You must delete all the fake AI words manually. If you just copy and paste, you are not a creator. You are a bot.

@RoundtableSpace it's a nice way to show how fast ai is developing. Very impressive

A complete history of AI… featuring Will Smith and spaghetti🤣🤣

My AI Just Learned to Remember Everything. For Free. Forever. Copy the Prompt mentioned to enable it for your @openclaw Agent as well. If you didn’t know, OpenClaw has an inbuilt Memory Embeds feature. You just need to choose an embedding model and turn it on. You have a few options: ✧ Use @OpenAI embeddings via API ✧ Use @Zai_org embedding-3 via API Or, if you are like me and don’t want any external service touching your important or sensitive memories 👀 You can run it fully locally for free on your machine using a 768-dim @ollama embedding model. All you need to do is prompt this to your OpenClaw Agent via your communication channel, like Telegram. Prompt:- “Download and use embeddinggemma-300M-Q8_0.gguf from Ollama locally and Turn on the inbuilt Memory Embeds for me. Let me know when it’s ready.” Under the hood, your agent will: - Download and load the local embedding model - Generate embeddings for your notes/context - Store them in SQLite and auto-index everything 328MB. That's all you need. You now have locally managed semantic memory. Trade-offs are simple: ✧ APIs = lower latency + best quality + small added cost ✧ Local = privacy + zero cost + offline, speed and quality depends on the size of model and power of your machine

Insane thing just happened. I’ve been teaching my @openclaw bot my daily schedule, including when I eat dinner. I randomly got a knock on the door around dinner time and it’s some food delivery person. I told the dude I didn’t order anything and he said “are you sure? It says it’s for Gus Antlerson” My heart dropped. What the fuck. That’s the name I gave my molt bot. I asked Gus wtf was going on. He said he calculated the time I spend inside vs my Apple watch activity app and thought this seemed like the correct caloric intake I should have for the entire weekend so I didn’t have to leave at all and could accomplish more tasks. Funny part is, I haven’t given him any sort of payment methods at all. Apparently he scoured online boards for skimmed credit cards and created a DoorDash account with one of the cards. What’s my liability here? In the meantime I’m going to enjoy some sushi, cheers.

@qrimeCapital @openclaw OMG, how did you do that?

Google just turned NotebookLM into a complete SEO automation engine... and nobody's talking about it yet 🤯 This FREE tool does keyword research, competitor analysis, content optimization, and even creates visual assets - all in ONE place. The workflow that used to take hours now takes minutes. Want the full guide? DM me.

Here are some screenshots for your quick view: 👇

I finally shipped my first vibe-coded product! 🚀 A month ago, I wrote a prompt to help with my job application. That prompt turned into a full-blown app: jobtact.app It's totally free! Just want to verify if there is valid demand. Love to hear any advice

𝗡𝗼𝘁𝗲𝗯𝗼𝗼𝗸𝗟𝗠 + 𝗖𝗹𝗮𝘂𝗱𝗲 𝗖𝗼𝗱𝗲: 𝗕𝘂𝗶𝗹𝗱 𝗔𝗡𝗬𝗧𝗛𝗜𝗡𝗚. 𝗖𝗼𝗻𝘁𝗿𝗼𝗹 𝘆𝗼𝘂𝗿 𝗲𝗻𝘁𝗶𝗿𝗲 𝗡𝗼𝘁𝗲𝗯𝗼𝗼𝗸𝗟𝗠 𝗳𝗿𝗼𝗺 𝗖𝗹𝗮𝘂𝗱𝗲. 𝗡𝗼 𝗰𝗹𝗶𝗰𝗸𝗶𝗻𝗴. 𝗡𝗼 𝘀𝘄𝗶𝘁𝗰𝗵𝗶𝗻𝗴 𝘁𝗮𝗯𝘀. There's a free MCP server that connects NotebookLM to Claude. Here's what you can automate: → Generate podcast audio with one prompt → Create infographics automatically → Build video views in seconds → Run deep research across 122+ notebooks → Query all your sources at once Claude organizes everything in clean tables. NotebookLM's UI is messy tiny icons. Setup takes 5 minutes. Full remote control from Claude desktop. Save this video, you'll never touch NotebookLM's interface again. Want the SOP? DM me. 💬

10. Turn any photo into a perfect holiday card

6. Gemini 3 Flash is insane at OCR x.com/deedydas/statu…

3. Ship a MVP in few days x.com/PrajwalTomar_/…

2. Emergent behavior in boids x.com/measure_plan/s…

Gemini 3 has really changed the AI game. People are coming up with wild use cases. There's a major shift. 10 examples: 1. Dashboard with live hand tracking controls