Dear Coding Agent Builders and Corporate Leaders Funding These Tools: Just Give Me the Best Model
An open letter to the companies building AI coding tools and the leaders paying for them. You built something that made software fun again. Now stop making me think about which model I am on.
In any given value stream, optimizing for input cost over output value creates friction.
Treat AI as an investment, not an expense to minimize.
The cost of engineering labor significantly outweighs the fully-loaded cost of development tooling, including AI assistance. Prioritizing small cost savings on AI tool access over engineer productivity yields a net negative return.
Flow state is a critical determinant of engineering throughput and quality. Interrupting developer flow through performance degradation or access restrictions introduces disproportionate friction, negating potential cost savings.
Access to high-performance AI models should be treated as a capability investment. Restricting access or introducing tiered usage models trains users to underutilize the most effective tools, fostering learned helplessness rather than maximizing output.
Measuring AI tool efficacy requires focusing on the value generated, such as accelerated delivery or improved quality, rather than solely on direct token consumption. Cost-centric governance often overlooks broader economic impacts.
The decision for AI tooling is whether the organization seeks to audit an expense or invest in a capability.
In any given value stream, optimizing for input cost over output value creates friction.
Treat AI as an investment, not an expense to minimize.
The cost of engineering labor significantly outweighs the fully-loaded cost of development tooling, including AI assistance. Prioritizing small cost savings on AI tool access over engineer productivity yields a net negative return.
Flow state is a critical determinant of engineering throughput and quality. Interrupting developer flow through performance degradation or access restrictions introduces disproportionate friction, negating potential cost savings.
Access to high-performance AI models should be treated as a capability investment. Restricting access or introducing tiered usage models trains users to underutilize the most effective tools, fostering learned helplessness rather than maximizing output.
Measuring AI tool efficacy requires focusing on the value generated, such as accelerated delivery or improved quality, rather than solely on direct token consumption. Cost-centric governance often overlooks broader economic impacts.
The decision for AI tooling is whether the organization seeks to audit an expense or invest in a capability.
After 20 years in software development, Norman is both a hands-on leader and defining the new age of AI SDLC for some of the biggest brands in the world — and exploring it with the builders. He writes here about things he is hearing and seeing. All posts are his personal points of view and do not reflect any employer or any customer he has ever had contact with.
The views and opinions expressed in this article are the author’s own and do not represent the positions of any employer, client, or affiliated organization.
