9 Comments

Great read! Ill be looking back at this when I start building!

Expand full comment

thanks Jon, don't forget to drop me a msg when you do!

Expand full comment

Thank you for sharing! I’m thinking of building something and the steps are of great help!

Expand full comment

glad this was useful for you! let me know if you need help with getting started!

Expand full comment

“But if you just rush into the building, you’ll loop yourself into problems because neither you nor the AI will have any priorities or any actual idea what a good result would look like.” —— 💯! Though I don’t come up with a plan all by myself, I ask Cursor to provide considerate structures and roadmaps, then we discuss all possibilities to decide on one approach 😊

Expand full comment

Exactly. Also, reasoning engines are fairly competent at sketching out basic architecture. You just give it the constraints and the tools of choice and it will find out how to make that happen. The main challenge is narrowing the scope. Feels like most models try to build corporate projects not MVPs :D

Expand full comment

That’s the thing!

Corporate projects might be feasible in the near future.

That said, I’m already amazed that with enough context, AI can uncover the most hidden bugs in enterprise code, something that would easily take me a full day!

Expand full comment

I'm waiting for the first AI agent that uses 3-4 reasoner models to sift through large-scale IT systems. I think we'll see the first one of these before Christmas

Expand full comment

The prediction is reasonable, and very exciting too!

Expand full comment