ITMan's Blog - 📚 Lifelong LearnerAmp Hack Night Singapore!
I just returned from the meetup and wanted to share a few thoughts while they’re still fresh.
To be honest, I came in expecting to see something new, maybe a beta feature, an experimental workflow, or a preview of what’s next. I was hoping for a live demo or a deeper look into what Amp is actively building.
Instead, the session felt more like a general meetup and open Q&A. There wasn’t a product demo or feature walk through, mostly discussion and sharing. That’s not necessarily bad, but it was different from what I anticipated.
That said, a few themes stood out.
1. Opinionated AI — Strong, but Closed
Amp takes a clearly opinionated approach. It decides which model provider to use for agents based on what it thinks is best for the user. This makes onboarding simple and reduces decision fatigue. However, I still feel there’s room for improvement.
There’s currently no way to control the model provider per agent. For developers who care about trade-offs (cost, latency, reasoning depth, privacy), this flexibility matters. An opinionated default is great, but optional control would make it stronger.
I wrote more about this earlier 👉 https://productsway.com/posts/power-up-1-amp-opinionated-ai-that-fits-real-developer-workflows
2. No Planning Mode
Another area that feels underdeveloped is planning.
If you’re building agents that require structured thinking, review loops, or multi-step reasoning, explicit planning becomes critical. Especially in ambient workflows, where tasks evolve over time.
I’d love to see:
Clear plan → execute → review phases
Visualized plans
Better support for structured feedback
This is something I’ve been exploring with tools like:
👉 https://productsway.com/posts/power-up-2-plannotator-visual-plan-review-for-coding-agents
Planning is where serious agent workflows will differentiate.
3. Mobile Is Coming
I heard that a mobile app is in progress. It wasn’t available to try yet, but that’s something I’ll be watching closely.
Conclusion
Overall, while I didn’t see the new beta or experimental feature I expected, the discussions were still valuable. It helped clarify where the product is strong , and where there’s opportunity to push further.