How Coders Can Survive in a The New ChatGPT World
Artificial intelligence, specifically generative AI fueled by large language models (LLMs), has the potential to disrupt the livelihoods of many coders. Nevertheless, some experts contend that AI will not immediately replace human programmers.
Tanishq Mathew Abraham, CEO of the medical AI research center MedARC and a recent Ph.D. in biomedical engineering from the University of California, Davis, advises coders to be wary of AI potentially replacing them. However, he believes that there are ways for software developers to remain relevant and indispensable in an era dominated by LLM-centered coding.
In preparation for the upcoming age of generative AI, here are several tips and techniques that coders can embrace to not only survive but thrive in this evolving landscape.
Stick to Basics and Best Practices
While the numerous AI-based coding assistants offer valuable support in code completion and generation, it is essential to recognize that the bedrock of programming still revolves around core skills. These skills include the capacity to comprehend and analyze one’s code as well as that of others, along with a profound understanding of how the code written aligns with and contributes to a larger system.
Find the Tool That Fits Your Needs
Selecting the appropriate AI-based tool is paramount. Each tool has its unique mode of interaction, and there are various ways to integrate them into your development workflow. Whether it involves automating the generation of unit tests, creating test data, or writing documentation, understanding how to effectively incorporate each tool can significantly enhance your productivity and efficiency.
Clear and Precise Conversations Are Crucial
When leveraging AI coding assistants, precision in articulating your requirements is crucial, treating it as an iterative process. Tanishq Mathew Abraham suggests providing explicit comments that precisely describe the desired code, enabling the assistant to generate relevant suggestions aligned with your needs.
For conversational AI programmers, proficiency in prompt engineering becomes essential. Framing prompts effectively is the key to obtaining accurate and desired responses from the AI model
Be Critical and Understand the Risks
Vaithilingam advises software engineers to exercise caution when dealing with the outputs of large language models, as these models often exhibit hallucinations and may generate code that is inaccurate or erroneous. Blindly relying on AI-generated code can lead to being trapped in an extensive debugging process, and detecting subtle bugs can prove to be a challenging task.
In an era dominated by generative AI, programmers must adapt to survive by embracing AI as a valuable tool and seamlessly integrating it into their workflow. This requires a clear understanding of the potential and constraints of these AI tools, all the while recognizing the importance of their human coding skills to thrive in this evolving landscape.
