Every few months, a new headline appears claiming that AI will soon replace software developers. The promise is seductive: lower costs, faster delivery, and fewer people required. But reality continues to tell a different story.
Recently, Ford Motor Company provided a powerful lesson in the limits of automation.
According to multiple reports, Ford invested heavily in AI-powered systems and automated quality processes. The expectation was that technology could help solve complex engineering challenges more efficiently and at lower cost. Instead, the company found that some of its most difficult quality problems persisted. The missing ingredient was not computing power. It was experience. [financialexpress.com], [forbes.com]
To address these issues, Ford reportedly hired and rehired hundreds of veteran engineers, many with decades of product and manufacturing knowledge. Their role was not only to solve problems directly but also to mentor younger staff and improve the AI systems themselves. [financialexpress.com], [techstartups.com]
This story highlights a truth that every software leader already understands.
Knowledge Is Not the Same as Experience
AI can consume vast amounts of information. It can generate code, explain frameworks, write tests, and even suggest architectures.
What it cannot reliably replicate is the judgement that comes from years of solving real-world problems.
Experienced developers recognise patterns that never appear in documentation. They sense when a seemingly elegant solution will become a maintenance nightmare. They spot hidden dependencies, anticipate edge cases, and understand the political, commercial, and operational realities behind technical decisions.
That knowledge is rarely written down. It is learned through successes, failures, outages, projects, customers, and time.
AI Is a Multiplier, not a Replacement
The most successful organisations are discovering that AI produces its best results when paired with experienced practitioners.
A junior developer using AI can produce more code.
An experienced developer using AI can produce better outcomes.
The difference matters.
Writing code has never been the hardest part of software engineering. Understanding the problem, choosing the right approach, balancing competing priorities, and managing risk are where real value is created.
Those responsibilities still belong to humans.
The Danger of Optimising for Cost Alone
When organisations view AI purely as a cost-cutting exercise, they risk removing the very people who make AI effective.
Ford’s reported experience suggests that experienced engineers were needed not just to solve problems but also to train, guide, and improve the automated systems themselves. [financialexpress.com], [forbes.com]
The lesson for technology leaders is clear: if you remove expertise, you also remove the context that makes AI useful.
The Future Is Human Plus AI
The debate was never really AI versus developers.
The future belongs to developers who know how to use AI effectively.
The companies that succeed will not be those that replace experience with automation. They will be those that combine decades of human expertise with the speed and scale that AI provides.
AI can write code.
Experience decides whether that code should exist in the first place.
Sources: Financial Express reporting on Ford’s rehiring of veteran engineers and Bloomberg-referenced executive comments. Forbes analysis based on Bloomberg reporting of Ford bringing back 350 veteran engineers. [financialexpress.com] [forbes.com]
