LLMs: Developer tools for success
With the rapid evolution of artificial intelligence and language models like GPT-4 and Copilot, there's a noticeable apprehension among developer. Some developers fear that these technological advances will negatively impact their skill level or that of their colleagues, while others are afraid of not keeping up with the curve or being replaced. While it's true that AI is changing the game, it's important to see these tools for what they are: allies that can help us improve.
The tool doesn't make the artisan
LLMs, code libraries, Stack Overflow... all these tools are valuable, but their effectiveness depends on two things:
- Understanding the tool: What is it meant for? What are its capabilities and limitations? This understanding is essential to know when to use one tool over another.
- Being aware of our mastery level with the tool: Am I using it correctly and to its full potential? Are the results I'm getting improving or stagnating? There are few substitutes for practical and concrete understanding of how to use a tool.
Let's take woodworking as an example. Giving a planer, a jointer, and a drill to someone with no woodworking experience won't make them a good carpenter. These tools, although powerful, require skills and expertise.
Conversely, an experienced carpenter will use these tools to accelerate their work and improve the quality of their creations. With the same equipment, an amateur and a professional won't achieve the same result. The bookshelf built by the professional will be more robust, more beautiful, and more durable than the amateur's. Even if both bookshelves can adequately store books, expertise and experience make all the difference.
The Value of Curiosity and Continuous Learning
As my colleague Joey Bolduc-Gilbert says, curiosity and the desire to understand are major assets. In the development field, these qualities allow us to stay current and stand out. Learning and adapting are essential in an constantly evolving environment.
LLMs can be seen as an extension of our capabilities. They help us solve problems, explore new approaches, and optimize our solutions. By saving us time on repetitive tasks, they allow us to focus on more complex and creative aspects of our work.
Instead of seeing LLMs as a threat, let's consider them as tools that amplify our skills. They can help us create prototypes quickly, explore different implementations, or grasp more complex concepts. Of course, they're not a magic solution, and without a solid knowledge base and good critical thinking, it's easy to misinterpret or misuse the information they provide.
A tool to enrich our work
AI doesn't replace human judgment, creativity, or deep understanding of complex systems. It's a tool that, when used properly, can help work faster and aid innovation. Using it or not is a choice, but choosing to do without it puts professional development at risk. Those who don't use it, or use it poorly, will inevitably fall behind while those who take the time to understand and master the tool will get ahead.
By cultivating curiosity about this new paradigm and constantly seeking to improve skills, LLMs can be used for what they truly are: tools designed to support and accelerate professional growth.
This article was first published in french on my LinkedIn profile before being translated to english using Claude 3.5 Sonnet and published here.