In a world run by increasingly powerful AI coding tools, creating software will become cheaper, leaving little room for traditional software companies. One analyst report states, “Vibe coding allows startups to replicate the functionality of complex SaaS platforms.”
Clues include wringing your hands and declaring that software companies are doomed.
Open source software projects that use agents to overcome long-standing resource constraints should logically be among the first to benefit from the era of cheap code. But that equation just doesn’t apply. In reality, the impact that AI coding tools have on open source software is much more complex.
According to industry experts, AI coding tools are creating as many problems as they’re solving. The easy-to-use and accessible nature of AI coding tools has created a flood of bad code that can overwhelm projects. Building new features has never been easier, but maintaining them is just as difficult, risking further fragmentation of the software ecosystem.
The result is a story more complex than simple software abundance. Perhaps it is too early to predict the imminent death of software engineers in this new AI era.
quality and quantity
Overall, projects that use open codebases notice a decrease in the average quality of their submissions. This may be a result of AI tools lowering the barrier to entry.
Jean-Baptiste Kempf, CEO of the VideoLAN Organization, which oversees VLC, said in a recent interview that “for people who aren’t familiar with the VLC codebase, the quality of merge requests we see is terrible.”
tech crunch event
boston, massachusetts
|
June 9, 2026
Kempf remains optimistic about AI coding tools overall, but says they’re best suited “for experienced developers.”
Blender, a 3D modeling tool that has been kept open source since 2002, had similar issues. Blender Foundation CEO Francesco Siddi said LLM-backed contributions typically “waste reviewers’ time and affect their motivation.” Blender is still developing official policies regarding AI coding tools, but Sidi said that AI coding tools are “not required or recommended for contributors or core developers.”
The flood of merge requests has gotten so bad that open source developers are building new tools to manage it.
Earlier this month, developer Mitchell Hashimoto launched a system that restricts contributions to GitHub to “verified” users, effectively shutting down the company’s open-door policy for open source software. As Hashimoto said in his announcement, “AI has eliminated the natural barrier to entry that makes OSS projects trustworthy by default.”
The same effect is being seen in bug bounty programs that open the door for outside researchers to report security vulnerabilities. Open source data transfer program cURL recently suspended its bug bounty program after being overwhelmed by what developer Daniel Stenberg described as “AI missteps.”
“Once upon a time, someone actually invested a lot of time into [in] “There was friction to begin with, but now there is no effort at all to do this,” Stenberg said at a recent briefing. The floodgates are open. ”
This is especially frustrating because many open source projects are also realizing the benefits of AI coding tools. Kempf says that building new modules for VLC would have been much easier if experienced developers were in charge.
“You can give a model the entire VLC codebase and say, ‘I’m going to port this to a new operating system,’” Kemp says. “Writing new code is convenient for advanced users, but difficult to manage for people who don’t know what they’re doing.”
competing priorities
A bigger problem with open source projects is a difference in priorities. While companies like Meta focus on new code and products, open source software work is more focused on stability.
“Big companies and open source projects have different issues,” Kemp commented. “They get promoted by writing code, not by maintaining it.”
AI coding tools are also emerging at a time when software in general is particularly fragmented.
Open source investor Konstantin Vinogradov says AI tools are entering a long-standing trend in open source engineering.
“On the one hand, we have an exponentially growing codebase with an exponentially growing number of interdependencies. And on the other hand, the number of active maintainers is growing, perhaps slowly, but clearly not keeping up,” Vinogradov said. “AI has accelerated both parts of this equation.”
This is a new way of thinking about the impact of AI on software engineering, and it has alarming implications for the industry as a whole.
If you think of engineering as the process of creating working software, AI coding makes it easier than ever. But if engineering is really the process of managing software complexity, AI coding tools can make it even more difficult. At the very least, it will take a lot of active planning and work to keep the sprawling complexity in check.
For Vinogradov, the result is a common situation in open source projects. There’s a lot of work to be done, but not enough good engineers to do it.
“AI does not increase the number of active and skilled maintainers,” he said. “It empowers good people, but all the fundamental problems remain.”
