As artificial intelligence continues to evolve, its impact on software development is becoming increasingly evident. According to Matt Garman, the new CEO of AWS, the rapid advancement of AI coding tools could fundamentally change the role of developers in the near future. Garman recently suggested that within the next two years, "most developers are not coding," a prediction that has sparked widespread discussion and concern in the tech community.
AI's Role in the Future of Development
During a recent conversation with AWS employees, Garman emphasized that the ability to write code is not the central skill that defines a good software developer. Instead, he argued that the core of a developer's role is the ability to innovate and build solutions that resonate with end users. He suggested that by 2025, the landscape of software development could be drastically different from what it was just a few years ago.
Garman’s comments echo sentiments expressed by other tech leaders, including Jensen Huang, CEO of Nvidia. At the World Government Summit in Dubai, Huang suggested that the future of programming could be one where coding is no longer a necessary skill, thanks to the advancements in AI. He later clarified at Nvidia’s GTC 2024 event that while many may not need to learn specific programming languages, there will still be a need for human programmers.
Not Everyone Agrees
Despite the bold predictions from some industry leaders, not everyone is convinced that AI coding tools will replace human developers anytime soon. Sid Sijbrandij, CEO of GitLab, argues that the conversation should shift from whether AI will replace developers to how AI can be leveraged to create real value in software development. He believes the focus should be on how businesses can harness AI to accelerate innovation and drive tangible outcomes for customers.
Similarly, Peter Schneider, senior product director at Qt Group, acknowledges the productivity gains provided by AI coding assistants, particularly in creating test cases and code documentation. However, he also notes that these tools are not yet sophisticated enough to replace human developers, especially when it comes to implementing industry-specific code. According to Schneider, the accuracy of AI-generated code needs to reach near-perfect levels before AI can substitute human programmers, and even then, human oversight will remain crucial.
The Need for Skilled Developers
Another expert, Peter van der Putten, director of the AI Lab at Pegasystems and assistant professor at Leiden University, also challenges the idea that AI will render traditional coding skills obsolete. Van der Putten warns that relying too heavily on AI-generated code without sufficient expertise could lead to significant technical debt and increased vulnerabilities within software systems. He argues that skilled developers will continue to play a critical role in ensuring the quality, safety, and performance of software.
"If you leave coding solely to AI tools, especially in the hands of non-coders, you risk generating unsafe, non-performant, or simply non-functional software," Van der Putten cautions. He believes that a strong understanding of major programming languages will remain essential for developers to manage and maintain the integrity of the software they create.
