Every few months, a new task management app is released, claiming it will change everything, but within days it quietly disappears.
There are already thousands of task management tools, and there will be thousands more. Yet no product has ever “killed the game.” That is because this space was never limited by tooling — it is limited by how humans actually form intentions and act on them.
Our takeaways from extensive user studies are:
AI shouldn’t just do things for us — it should think with us by acting as a partner in reflection and planning (TaskWise)
The real opportunity is designing AI that creates just enough participation to help people turn vague intent into actionable context — not replacing thinking, but shaping it at the right moment. (Kairotask)
So we built a smart task management tool that continuously learns users’ daily habits in the background and provides more intelligent reminders. It also enables collaboration between users’ personal AI assistants, so users only need to focus on their own tasks while the assistants can handle most work-related social interactions automatically.
We believe this idea has strong potential, and we have already discussed it with several VCs from Sequoia Capital, MiraclePlus, and others. They also believe the direction is promising (trust me).
However, something is holding me back right now. Applications such as OpenClaw and similar AI systems seem capable of accomplishing this kind of work almost instantly. It makes me wonder whether what we are building still has real meaning.
After seeing GPT-Realtime-2 and the Interaction Model from TML, I started to think that they may provide a completely new perspective for task management and AI assistants. Perhaps this could push what we want to build one step further.
We are still building this.
If you are interested, feel free to contact me by email.