Tokenomics
What I often see – in blogs as well as LinkedIn – is that exceptional developers like Daniel Terhorst-North spend inordinate amount of time wrangling agents (to use Simon Wardley’s wording) to get AI to produce quality software on a consistent basis.
Is this really cost effective in the long run?
Tokenomics is finally taking hold and the “all-you-can-eat” buffet has shut down. And it has shut down without an established recognized standardized integration of AI into large corporations, that need to coordinate their efforts over a hundred teams. What works for your weekend project is not enterprise-grade software development.
A little bit like scrum – I don’t see how this scales. I don’t think it adds up – you can’t just create one set of guardrails and call it a day. The models keep changing as does any development environment. How do you stay on top of that, when you’re not even coding anymore?
Am I missing something?