Just curious… I built an ai receptionist using Gemini 3.1 flash and Livekit and doing great for my clients. Question is I’ve been ask a few times about an outbound receptionist so. Wondering if anyone is using this for outbound?

Both Claude Code and Codex can run on the same project. You just need to know what each one looks for so you don't have to duplicate files or rebuild your setup. This video walks through what to swap, what stays the same, and a quick prompt you can use to convert any Claude Code project so Codex can read it too.

I am working on building the Hermes Agent(s) step by step. I just have one question (I'm not a coder). In his other video I built my "first agentic agent." In my visual studio code I have all those docs and folders up in my explorer (see attached screenshot). Do I need to add a new folder or start fresh or anything in order to build the hermes agent and keep everything organized? Cause my explorer obviously does not look like his from the video. I'm worried that I'm going to use up a crap ton of tokens building these agents if I keep my explorer as is because it will index through all the other stuff I built, right?

Same prompt. Four different results I ran the exact same northern lights prompt through 4 different AI tools and the differences are immediately obvious — once you know what to look for The four tools used: — Nano Banana Pro — Nano Banana 2 — Seedream 4.5 — ChatGPT Image 2 Your job is simple Look at the 4 images. Match each one to the tool that made it. Drop your answer in the comments like this: Image 1 — Nano Banana Pro Image 2 — ChatGPT Image 2 Image 3 — Seedream 4.5 Image 4 — Nano Banana 2 No guessing in silence. Commit to your answer. Once enough people have guessed I'll reveal the correct order and break down exactly what gave each tool away — the way it handled the aurora glow, the ice reflection, the star field, the atmosphere Drop your answer below 👇

I’ve been building automation systems… but clients actually buy emotional relief 😅 Nobody wakes up thinking: “Wow I really need an AI workflow.” They think: - “Why are leads slipping through?” - “Why is my team slow to reply?” - “Why does everything feel chaotic?” That changed how I rewrote the offer today. Before: “AI-powered lead automation system” Now: “Never lose a warm lead because your business replied too late.” Same system. Completely different feeling. Also started testing response timing. A lead getting a useful reply in 2 minutes vs 2 hours feels INSANELY different psychologically. The crazy part? Most businesses still operate like it’s 2018 when it comes to speed. Feels like there’s a huge opportunity just in fixing responsiveness.