Not because it was bad. Because it simply didn't fit the way you work. I think we're at a point where there are new AI tools launching almost every day. Some genuinely improve how we build. Others are exciting for a week... then quietly disappear from our workflow. For me, the biggest difference isn't whether a tool is "the best." It's whether it consistently helps solve real problems. I'm interested to hear from people actually building things. What's one AI tool you tried, thought would become part of your daily workflow, but eventually stopped using? And what made you move away from it?

Hi all. Most of you know the AIOS concept, so I'll skip the basics and get into what I built and where I want your feedback. It's heavily inspired by ICM, OKF, and the LLM Wiki work. The shared idea across all three: keep the business as plain, structured, human-readable files (markdown), not a vector DB, so any model can load and reason over the whole thing directly, and so the owner keeps their system instead of renting it. Everything else follows from that. It's your whole business (context, clients or customers, services, workflows, knowledge) as one open, file-based system that any AI can run. You download it and own it, no subscription, provider-agnostic. The autonomous agent is the piece I care most about, and it does two jobs: - Onboarding: a conversation that builds your structure and context as you talk. It replaces the slow manual interview, and the familiarity you build during setup carries straight into using it. - Ongoing assistant: the same agent teaches you the system, builds and updates your skills, runs skills on demand or on a schedule, and pulls from connected tools. Text-first throughout. Some design decisions: - Skills you build and run, not a fixed roster of agents. Easier to grow, version, and maintain. - A shared engine plus per-vertical content packs, so the same system adapts to different businesses without forking. Provider-agnostic model routing, so you can point it at whatever model you want. - Where it's at. I run my marketing agency on it, and I'm now offering it to owners in the 4 verticals I know best (agencies, home services, ecommerce, personal services). I'm actively improving it as I learn from each build. What I'd love feedback on: - The file-native, no-vector-DB bet. Where does it stop scaling? Agent-led onboarding as a pattern. Better or worse than a form plus templates? - Skills-as-capability vs an agent swarm for the ongoing assistant. What would you need to see to trust this to run real parts of your business?

Hey everyone, It's been a few months since I've posted here. During that time, I've been doing a lot of thinking about where I actually want my career to go. I got into the AI space because of how exciting everything was and, honestly, because of all the hype. But over the last few months, I've realized something about myself: I don't get the same excitement from AI as I do from writing software. Whenever I'm coding, I completely lose track of time. Solving problems, building applications, and writing clean code genuinely makes me happy. It's one of the few things that gives me that feeling of wanting to keep going for hours. I think I reached a point where I felt overwhelmed trying to keep up with how fast AI is moving. Instead of enjoying the process, I started feeling like I was constantly behind, and eventually I stepped away to focus more on school. Now I'm wondering if this is a sign that I should lean into software engineering instead of trying to force myself into AI. For those of you who've been in tech longer: - Have you ever realized you were pursuing something because of the hype rather than genuine interest? - If you discovered you enjoyed software engineering more than AI, would you fully commit to it or keep one foot in both worlds? - Would you pivot your personal brand, including LinkedIn, or would you keep AI as part of your profile? I'd really appreciate hearing your experiences. Thanks in advance! (Though am still a cadet and it's my major in the university)