Artificial intelligence (AI) has revolutionized the way we work, especially in technical fields like software development. Tools such as ChatGPT, GitHub Copilot, and CodeiumAI are making coding more efficient by offering solutions for everything from automated log analysis to code completion. In this article, We’ll share our personal experiences with these tools, exploring their unique capabilities, differences, and how they help us streamline our workflows while staying mindful of sensitive information.

#1 The versatile ChatGPT

‍

“ChatGPT has proven to be an effective tool in my daily work in many areas, such as automated analysis of huge logs, faster searching in C++, text generation and in certain cases, to support code review.”

‍

First introduced in 2018, ChatGPT has evolved through several versions. The difference between GPT-4 and GPT-4o lies in their optimization and capabilities. GPT-4o is a more advanced version, offering enhanced multimodal abilities within a single model.

‍

#2 Using ChatGPT to obtain specific information

‍

“I ask ChatGPT several times daily to provide information about APIs and industry recommendations. I always take care not to provide sensitive information in any conversation. I consider this a key usage point.”

‍

#3 Improving code efficiency with CodeiumAI

‍

“I started using AI a few years ago. I beta tested GitHub Copilot, but mainly use CodeiumAI.”

‍

‍GitHub Copilot is a paid AI code assistant that provides real-time code suggestion within GitHub-integrated environments.

‍CodeiumAI is a free AI code assistant offering multi-platform support and focuses on code completions and refactoring.

‍

#4 Talking with Claude

‍

“I talk to Claude daily about general coding issues and solving specific problems. In addition, it generates RegEx and CSS well, with only a little fine-tuning needed.

Claude's development company has introduced Artifacts, which is AI-generated interactive output, e.g. program code, block diagrams, or even flowcharts. I find this particularly useful.”

‍

‍Claude is an AI by Anthropic that helps with tasks involving understanding and generating human language (like answering questions or summarizing text). It supports coding by providing code completions, debugging, and optimization suggestions based on context.

‍

#4 GitHub CoPilot to support coding

‍

I've been testing CoPilot for a month. Its effectiveness varies, but I've found that a lot depends on the clarity of the prompts. CoPilot assists me in writing code, sometimes completing 2-3 lines for me, which I then check. There have been cases where I could enhance the code it provides. The chat feature is also impressive, faster and more specific than a simple search.

‍

REMEMBER! Always take extra care to protect sensitive information when working with AI.

‍

‍