looking to connect with people building cool stuff:
• micro-saas
• AI agents
• indie apps
• dev tools
• GTM systems
• building in public
what are you working on? 🤝
@dogramaci probs a wrong customer facing action. you can roll back most system changes, but rebuilding trust is much harder. that’s where i’d keep a human approval step.
@i_mika_el x-builder sounds useful. right now i’m figuring out a simple content-to-demo loop for Spinnable, but still early.
what made you build it?
Let's connect!
OpenAI says Astra can find unknown security flaws and develop exploits with the right tools and access.
So here’s the 2026 operator test:
Before giving an AI agent production access, always ______.
@BernieSanders A global pause sounds simple. Enforcement is the hard part.
What capability triggers it? Who audits the labs? How is compliance verified? What stops bad actors from continuing in secret?
Without answers, a pause is a slogan, not a safety system.
Building in AI, SaaS, or startups?
I’m looking to connect with more people who are building, shipping, and figuring things out in public.
Reply with what you’re working on. 👋
AI agents are easy to trust when they only suggest.
The hard part starts when they can act.
What should an AI worker do autonomously, and what should always require human approval?