← Back to Blog

I Have the Most Efficient AI Partner Possible. I Still Go Slow.

By Jamie Williams • December 28, 2025

Hot take: Even with an efficient AI that knows you, work doesn't become instant.

I've spent months building a collaboration system with my AI. We have shared context. Communication shortcuts. It knows my working style, my projects, my preferences.

And you know what? We still go slow. Have I been able to build stuff fast? Yes. It's incredible. No denying that, but there is still down time.


What doesn't disappear

  • Discovery time. You still need to figure out what to build before you build it.
  • Brainstorming. Ideas need space to breathe. AI helps, but it doesn't replace the wandering.
  • Edits. First drafts are fast. Getting it right still takes iteration.
  • Thinking. Some things just need time to marinate.
  • Incorporating Feedback You still have to put the consumer at the heart of your product. And get real feedback from users.

The shape of work changes. The existence of work doesn't.


Maybe the old frameworks are obsolete

Two-week sprints. Quarterly planning. Stand-ups. Retros.

I'm not sure these make sense anymore when you can prototype something in an afternoon.

I've been experimenting with something different: session sprints. One session, one focus, ship or learn something. Sometimes it makes more sense to build the thing first, then get feedback. Planning on the fly for some small pieces of work. Responsive instead of rigid.

It's messier. It's also more honest about how creative work actually happens.


The vision I'm working toward

Four-hour workdays. Two-hour focused sprints. Real partnership with AI, not just delegation.

Not because AI does everything. Because the leverage means you can do meaningful work without grinding for eight hours.

We're not there yet. But I'm closer than I was six months ago. I love that with AI we can move faster with intention.


Efficient ≠ instant

Here's what I want people to understand: AI makes you faster. It doesn't make you a machine.

You'll still have tech debt. You'll still need discovery time. You'll still iterate.

And honestly? That's fine. The human parts of work—the thinking, the wandering, the "what if"—those are the good parts anyway. And you still need consumer feedback. QA becomes more important. But now you can test out multiple concepts and build things quickly to get feedback quicker. And that is very fun for a product person like me. I'm excited to see how the paradigms of product work shift and what this means for traditional product management build cycles.


The goal isn't to eliminate work. It's to make the work go faster, so you can build better. And have more time to focus on work that's actually worth doing.

📝 Note: Ideas and opinions are mine, but this post may have been written with AI assistance. Please note mistakes can happen. This is for general information and entertainment purposes, not a substitute for professional advice (e.g., medical, legal, financial). Use at your own risk. Opinions expressed are my own and do not necessarily reflect the views of any organizations, employers, or affiliates I may be associated with.

Product ManagementAI Strategy
J

Jamie Williams

Product leader and builder at the intersection of AI, data, and culture. Based in Cincinnati. Shipping products, testing ideas, writing about tech that actually works.