AI tools are revolutionizing software development by automating repetitive tasks, refactoring large code, and identifying bugs in real time. Developers can now generate well-structured code from plain language prompts, saving hours of manual effort. These tools learn from large codebases, providing contextual recommendations that improve productivity and reduce errors. Rather than starting from scratch, engineers can prototype quickly, iterate faster, and focus on solving increasingly complex problems.
As code generation tools gain popularity, they raise questions about the future size and structure of engineering teams. Earlier this year, Garry Tan, CEO of startup accelerator Y Combinator, noted that about a quarter of his current clients use AI to write 95% or more of their software. In an interview with CNBCTan said: “What this means for founders is you don’t need a team of 50 or 100 engineers, you don’t need to raise that much. The capital lasts a lot longer.”
AI-powered coding can offer a quick solution to businesses under budget pressure – but its long-term effects on the ground and labor pool cannot be ignored.
As AI-powered coding grows, human expertise could decline
In the age of AI, the traditional path to coding expertise that has long supported senior developers could be under threat. Easy access to Large Language Models (LLM) allows beginning coders to quickly identify code problems. While this speeds up software development, it can distract developers from their own work, delaying the development of basic problem-solving skills. As a result, they can avoid the concentrated, sometimes uncomfortable hours required to gain expertise and progress on the path to senior developer success.
Consider Anthropic’s Claude Code, a terminal-based assistant built on the Claude 3.7 Sonnet model, which automates bug detection and resolution, test creation, and code refactoring. Using natural language commands, it reduces repetitive manual work and increases productivity.
Microsoft has also released two open source frameworks – AutoGen and Semantic Kernel – to support the development of agentic AI systems. AutoGen enables asynchronous messaging, modular components, and distributed agent collaboration to create complex workflows with minimal human intervention. Semantic Kernel is an SDK that integrates LLMs with languages like C#, Python, and Java, allowing developers to create AI agents to automate tasks and manage enterprise applications.
The increasing availability of these tools from Anthropic, Microsoft, and others may reduce opportunities for coders to refine and deepen their skills. Rather than “banging their head against the wall” to debug a few lines or select a library to unlock new features, junior developers can simply turn to AI for help. This means that experienced coders with problem-solving skills honed over decades could become an endangered species.
Over-reliance on AI to write code risks weakening developers’ hands-on experience and understanding of key programming concepts. Without regular practice, they may struggle to debug, optimize, or design systems independently. Ultimately, this erosion of skills can undermine critical thinking, creativity, and adaptability – qualities essential not only for coding, but also for evaluating the quality and logic of AI-generated solutions.
AI as a Mentor: Turning Code Automation into Practical Learning
While concerns about AI diminishing the skills of human developers are valid, businesses should not ignore AI-supported coding. They just need to think carefully about when and how to deploy AI tools under development. These tools can be more than productivity boosters; they can act as interactive mentors, guiding coders in real time with explanations, alternatives, and best practices.
When yoused as a training tool, AI can reinforce learning by showing coders why the code is broken and how to fix it, rather than just applying a solution. For example, a junior developer using Claude Code can receive immediate feedback on ineffective syntax or logic errors, as well as suggestions linked to detailed explanations. This allows for active learning and not passive correction. It’s a win-win: speed up project deadlines without doing all the work for beginner coders.
Additionally, coding frameworks can support experimentation by allowing developers to prototype agent workflows or integrate LLMs without the need for expert-level prior knowledge. By observing how AI creates and refines code, junior developers who actively engage with these tools can internalize patterns, architectural decisions, and debugging strategies, mirroring the traditional learning process through trial and error, code reviews, and mentoring.
However, AI coding assistants should not replace real mentoring or pair programming. Pull requests and formal code reviews remain essential to guide new, less experienced team members. We are far from the point where AI alone can improve the skills of a junior developer.
Businesses and educators can build development programs structured around these tools that emphasize code understanding to ensure AI is used as a training partner rather than a crutch. This encourages coders to question AI results and requires manual refactoring exercises. In this way, AI becomes less of a substitute for human ingenuity and more of an enabler of accelerated experiential learning.
Bridging the gap between automation and education
When used intentionally, AI doesn’t just write code; he teaches coding, combining automation and education to prepare developers for a future where deep understanding and adaptability remain essential.
By embracing AI as a mentor, as a programming partner, and as a team of developers that we can direct toward the problem at hand, we can bridge the gap between effective automation and education. We can enable developers to develop alongside the tools they use. We can ensure that as AI evolves, human skills also evolve, fostering a generation of coders who are both efficient and highly competent.
Richard Sonnenblick is chief data scientist at Plan view.
