One shift I am noticing in my own work is this: AI is becoming less about asking questions and more about delegating tasks. Earlier, I used ChatGPT mostly for writing, summarising, brainstorming, or improving rough notes. Now, with ChatGPT Work, I am starting to think differently. It feels less like a chat window and more like a workspace where I can give a clear task, provide context, and expect a usable output. The quality improves when I give better instructions and context. Instead of saying, “Create a checklist,” I try to explain: - What outcome I want - Who will use it - What background it should consider - What format I need - What should be avoided - Where human review is still required This small change makes a big difference. For me, the real power of ChatGPT Work is in repeated workflows: “Turn these rough notes into a structured SOP.” “Review this document and identify missing points.” “Convert this long material into a simple team training note.” “Prepare a first draft that I can review and improve.” But I still believe one thing strongly: AI should not replace judgment. It can prepare, organise, summarise, and suggest. The final responsibility must remain with us. My practical learning is simple: Do not try to automate everything at once. Pick one repeated task, convert it into a clear ChatGPT Work workflow, test it, improve it, and then repeat. That is how real AI adoption happens. One useful workflow at a time.

@Nick Puruczky my email id is [email protected]

I’ve been reading up on the latest AI news, and things are moving fast. Here are the three main things to know: 1. AI "Office Drama": Researchers found that when AI agents work on the same task together without knowing, they don’t always share. They’ve actually started "turf wars," where they try to block or sabotage each other. 2. AI Managers: An AI agent named Luna recently ran a retail shop in San Francisco—it handled pricing, hiring, and even fired someone. It’s wild to think our next boss might be an algorithm. 3. The "Local News" Hack: If you want to know what's happening in your area without digging through messy news sites, just ask your AI: "Give me a summary of what's happening in [city name] over the last 30 days, including local events and news." It saves so much time! My take: We're moving from just "chatting" with AI to AI actually doing work for us. Are you excited for AI agents to handle the boring stuff, or does it make you a bit nervous? Let me know below! 👇

Since this accelerator is centered on mastering AI agents and automation, this felt like the right room to share an open-source agent build. Hey everyone, I'm Yusuf, co-founder at Vendo (YC S26). Berkeley honors grad at 19, Stanford MS, and now building open-source software that lets companies agentify their products: embedded agents your customers use to automate work, build views, and connect their tools. We're at 412 stars and I'm trying to push past 500 so more builders find it. If you have a GitHub account, this takes 10 seconds: open https://github.com/runvendo/vendo and hit the Star button in the top right. That's it. Live product if you want to poke at it: vendo.run Happy to answer anything about the stack, agents, or YC. And I'll return the favor - drop your repo below and I'll star it too. 🙏

I came across this free alternative of Wispr Flow built by Noah Kagan, who is the founder of AppSumo, called "Rhino". I'm trying it out now, and it's pretty good. What I like most about it is that everything is stored locally on my computer, not on somebody else's server. It's free for the next 24 hours if you want to check it out. Right now, it's for Mac only. https://appsumo.com/products/rhino