CoPilot and ChatGPT for Software Development: Insights from Developers at LegalMatch

In today’s world, AI-driven software development has evolved from a mere futuristic concept to a practical reality. With over 1.2 million software developers worldwide utilizing AI tools to streamline their coding process and enhance efficiency, the potential for time and cost savings is huge. 

At LegalMatch, our developers have enthusiastically embraced AI tools, recognizing the significant advantages they bring. These tools have proven to be valuable assets by simplifying tasks and enhancing productivity for our team. However, it’s worth noting that while AI tools offer undeniable benefits for others, some developers express a degree of caution, acknowledging that they cannot solely rely on AI for all aspects of their work. 

Two of the most frequently used AI tools used by our developers to improve their workflow are CoPilot and ChatGPT.  Now, to gain further insights into our developers’ experiences with these tools, we conducted a mini survey. Here’s what some of them had to share about their encounters with these tools: 

 

CoPilot 

CoPilot is an AI-powered code completion tool developed by GitHub in collaboration with OpenAI. It integrates with popular code editors, such as Visual Studio Code, and provides real-time suggestions and autocompletion for code snippets while developers are writing software. 

 

“It generates a lot of boilerplate code that I would need to write by myself otherwise. Sometimes it can just generate some tricky functions that I simply describe in a comment, then press “Enter” and already get some workable output. That’s especially useful with languages when I’m not really an expert, like Groovy or Terraform.” Nikolai P., Senior Software Developer 

 

“Sometimes we encounter weird bugs that require time to investigate and resolve. And if the investigation doesn’t go well enough, our usual approach would be to try to google similar symptoms or submit a question to Stack Overflow. Now with the AI assistant, we are able to submit the question in natural human language, no matter how silly it sounds, and immediately receive some valuable clues.” Nikolai P., Senior Software Developer 

 

“Initially, I was hesitant about the idea of having an AI companion for coding, but exploring GitHub CoPilot totally changed my perspective. Yes, it looks scary that AI knows more language and code, but it will not work without a dev driving the coding. If you’re the type of developer who writes your algorithm as a TODO for a guide to complete your algorithm, CoPilot will greatly change your game as you just have to convert that TODO into a comment and you’ll have like multiple people interpreting, and coding the comment giving different approaches. This tool will greatly help a developer be more productive. 

 There are still many things to explore about CoPilot, so I’ll continue using the tool and add the other good things later.” – Alelie S., Senior Java Developer

 

Chat GPT 

Chat GPT (Generative Pre-trained Transformer) is an AI language model developed by OpenAI which offers developers a powerful way to incorporate natural language understanding and generation capabilities into their software, enhancing the conversational aspects of applications and systems. 

 

“It is still not a silver bullet. It is only useful for entirely new codes that have fewer dependencies on our legacy/existing codes because ChatGPT and other AI researcher tools have no access to our private code repositories. It will take a good developer to identify how to incorporate the generated code into our existing codebase. 

 Basically, we really don’t know if it’s really AI in the background. For normal end users, how can they tell between a machine-learning chatbot and an ordinary chatbot?  AI has a wide scope of definition.” Ron K., Senior Java Developer 

 

“ChatGPT has made my life easier and more effective. They cover things I might miss and enhance my development process. With our knowledge combined with AI, we’re on the right track. Welcome to the future!” – Orjan B., Software Engineer 

 

While AI tools are undoubtedly capable of saving time and enhancing productivity, they are not intended to replace a developer’s expertise or critical thinking. Instead, they should be employed as a valuable assistant to a developer’s skills and judgment. 

 

 

Leave a Reply

Your email address will not be published. Required fields are marked *

50 − = 47