Paul Ford: AI made bad code easy, not good software optional
Simon Willison's blog posted a short quote from Paul Ford's article "A.I. Was Supposed to Give Us New Killer Apps. What Happened?" Ford writes that it once looked like software developer jobs like his were finished, since there seemed to be no way to compete with tireless AI systems. He says the industry is now realizing the opposite: making truly cutting-edge software still requires humans to think and work together, combining their skills and practicing their individual crafts. In his view, AI can write very good software, but it just as easily lets someone do someone else's job badly, which he ties to why many AI-built software projects fail. He sums up the shift with the line that now that everyone can code, it has become clearer why many people should not. Willison's post gives only the quote and its source line, without naming the publication that ran Ford's original piece or describing Ford's professional background.
Key facts
- Simon Willison's blog quoted a passage from Paul Ford's article "A.I. Was Supposed to Give Us New Killer Apps. What Happened?"
- Ford says it once seemed AI would make developer roles like his obsolete, since tireless AI systems looked unbeatable
- He argues the industry is realizing that cutting-edge software still requires humans thinking and working together, using their skills and practicing their crafts
- Ford says AI can write very good software but also makes it easy to do someone else's job badly, which he links to why many projects fail
- His closing line: now that everyone can code, it is clearer why many should not
Why it matters
The quote pushes back on the idea that AI coding tools are close to replacing software developers outright. Ford's argument is that the bottleneck was never the ability to produce code, since AI has made that trivial, but the judgment and collaboration needed to turn code into cutting-edge software. That reframes AI coding progress: writing code got easier, but writing good software did not.
Who it affects
The point is aimed at software developers and the teams and companies building products with AI coding assistance. Ford frames it as a shift within his own industry, describing developer roles like his as the ones that briefly looked threatened.
How to use it
There is no product, price or tool to adopt here. The practical takeaway from Ford's argument, as quoted, is a caution against assuming AI-generated code removes the need for experienced developers and teamwork on any given project.
How solid is it
The claim is a personal argument from Paul Ford's article, relayed as a short quote on Simon Willison's blog rather than reported with data. No figures, studies or named project failures back the claim that 'all those projects fail'; it rests on Ford's own read of the industry.
Risks and caveats
The source gives only this excerpt and its title line: the outlet that published Ford's original article, his professional role and when the original piece ran are not stated. The 'projects fail' claim is asserted, not demonstrated, and should be read as one writer's opinion rather than an established finding.
“For a while, I must admit, it looked as if software developer roles like mine were done for. How could we fight against tireless robots? But our industry is slowly realizing that making truly cutting-edge software still requires humans to think and work together, to maximize their skill sets and to practice their respective crafts. A.I. can write very good software, but it also makes it easy to do someone else's job badly, which is part of why all those projects fail. Now that everyone can code, it's become clearer why many shouldn't.”
— Paul Ford, "A.I. Was Supposed to Give Us New Killer Apps. What Happened?"