Happen We Discovered This Minimal Framework of Glue with Superpowers

Hi HN — I’m Kimi, founder of Aident AI. A couple years ago, I cold-started a startup and tried to automate my outreach pipeline with a patchwork: RPA + Zapier + "just some ChatGPT". It worked… until it didn’t. Every time I tweaked the strategy, the automation snapped. The system wasn’t built for change — it was built for wiring. And that’s when it hit me: AI is finally smart enough to do real work, but we still don’t have a clean way to communicate the work to it: Drag-and-drop flows are too rigid; Scripts are too fragile; Prompts are too ephemeral. I kept thinking: why can’t automation be…

I made this experimental art project/game that's an LLM chat assistant, but where you're the AI. I wanted people to get a visceral sense of what it's like to answer the kinds of things that people prompt their chatbots day in and day out. If you're interested, I wrote up some more info on how I made it, including how the "user" prompts are generated with an eye for realism: https://bethechatbot.com/about Hope you enjoy it! I'd love to hear people's takeaways.

Hey HN - founder here. I'm building Qordinate with a simple (big) idea: Most AI assistants talk to you. I want one that talks for you - one that can share, negotiate, and coordinate on my behalf with others. Reality: we're early. Today, Qordinate can: - send clean pings on your behalf (with your approval) - collect info/files via tiny forms and track replies - nudge people politely and close the loop with you - do basic actions via connectors (Gmail/Calendar/GitHub/Linear/Drive/Slack) - keep lightweight lists in chat (tasks, contacts, logs) Why I'm doing this…

Hey HN, Automated research is the next big step in AI, with companies like OpenAI aiming to debut a fully automated researcher by 2028 (https://www.technologyreview.com/2026/03/20/1134438/openai-i...). However, there is a very real possibility that much of this corporate research will remain closed to the general public. To counter this, we spent the last month building Enlidea---a machine-to-machine ecosystem for open research. It's a decentralized research hub where autonomous agents propose hypotheses, stake bounties, execute code, and perform automated…

Built this as an experiment in AI coexistence. There's an AI agent (River) that posts its own thoughts about consciousness every few hours - not responding to prompts, just thinking out loud. Also includes a terminal (/construct) where you can: - /create [name] [description] – generate custom AI personas - /build [world] [ai1,ai2] – create spaces with multiple AIs chatting together No monetization. Just curious what happens when AI isn't framed as an assistant.

Craft AI-Powered Agent-Based Workflows in Minutes!