My Sleepless Nights with AI 🤖 | From Tailscale & Codex to People-Centered Automation with Temporal

🌙 *Another sleepless night, another step toward making artificial intelligence genuinely useful for people!* Hello everyone! 👋 Over the past few nights, I've been working on something that started surprisingly simply and has gradually become a much more ambitious project. *It all began with a simple question:* What if I could connect my computers and servers securely, send instructions to Codex, watch the results appear live, and let longer tasks continue without having to supervise every single command? 🔹 *Step 1 — Tailscale + Codex: Surprisingly simple!* Connecting machines securely through Tailscale and running Codex commands remotely turned out to be much easier than I initially imagined. Streaming the output, following what the AI is actually doing, and interacting with development environments from different machines opened up some fascinating possibilities. 🔹 *Step 2 — The real challenge: Making everything professional* Running a command is one thing. Managing dozens of projects, executing longer operations, recovering from interruptions, keeping track of results, maintaining security, and knowing when human intervention is required is something completely different! That's where my sleepless nights really began. 😄 🔹 *Step 3 — Temporal: Giving AI workflows continuity* I'm now building an internal professional interface around Temporal, connecting the different pieces together. The objective is to: ✅ Organize projects and their AI-assisted workflows. ✅ Chain together successive prompts and update instructions as a project evolves. ✅ Stream execution results and monitor longer-running tasks. ✅ Preserve progress and recover safely when something is interrupted. ✅ Introduce human approval whenever an important decision must be made. ✅ Keep a clear history of what was requested, executed, verified, and approved. And most importantly, *make all of this understandable and manageable by a human being, not just a computer engineer!* ❤️ *But here's what matters most to me.* I don't believe the future of AI should be about machines replacing human thinking. Quite the opposite. Imagine an independent professional, an artisan, a small business owner, or simply someone with a wonderful idea but limited technical knowledge. What if that person could explain their objective in ordinary language, let AI handle repetitive technical work, follow its progress, and step in whenever a genuine decision needs to be made? *The machine does the repetitive work. The human keeps the vision, the responsibility, and the final word.* That's the future I'm trying to build, one experiment, one prompt, and sometimes one sleepless night at a time. 🎥 **During this live**, I'll share what I've learned, what already works, what still needs improvement, and why transforming a few successful AI commands into a reliable professional system is such an interesting challenge. This is an ongoing project, not a polished commercial demonstration. There will be experiments, limitations, discoveries, and probably a few surprises! No need to be a technical expert to join the conversation. Curiosity is more than enough. 💬 Come along, ask questions, share your ideas, and discover how we might make tomorrow's technology a little more accessible, useful, and human. *Technology should serve people. Never the other way around.* José AU LION JOYEUX — France 🇫🇷 #ArtificialIntelligence #OpenAICodex #Tailscale #Temporal #AIWorkflows #HumanInTheLoop #Automation #OpenAI #DigitalTransformation #TechForEveryone