AI Tooling for Software Engineers in 2024: The State of the Art

As we reach the August of 2024, it's clear that AI tools have become an integral part of many software engineers' workflows. At NullPointerClub, we crawled the internet to understand how developers are using these tools and what impact they're having on productivity and code quality. Here are the key findings:

🏆 GitHub Copilot and ChatGPT Dominate

Our data revealed that GitHub Copilot and ChatGPT are by far the most popular AI tools among software engineers. In fact, as many professionals are using both of these tools as all other AI tools combined! Other notable tools include:

• Google's Gemini

• Anthropic's Claude

• JetBrains AI

• Tab9

• Cursor

• Codeium

🛠️ Common AI-Assisted Workflows

Developers who have used AI tools for more than six months about their typical workflows. Some common patterns emerged:

1. Using Copilot directly in the IDE for code completion and generation

2. Chatting with AI bots like ChatGPT instead of googling for answers

3. AI-assisted debugging to identify and fix issues faster

4. Using AI to quickly learn unfamiliar languages and frameworks

One data engineer shared: "I use GitHub Copilot for daily coding and ChatGPT 4 for complex, open-ended design discussions."

✅ The Benefits of AI Tools

When AI tools work well, they can have a massive positive impact on developer productivity. Respondents highlighted benefits like:

• Completing projects faster

• Increasing test coverage

• Making experimentation and prototyping easier

• Onboarding new team members more quickly

❌ The Challenges and Pitfalls

Of course, AI tools are not without their drawbacks. Some of the most common complaints include:

• Poor or incorrect output

• AI "hallucinations" (making things up)

• Developers over-trusting the AI and not validating its suggestions

As one senior software engineer put it: "My favorite is GitHub Copilot. It's the least bad of all the AI tools I have access to."

📈 The Future of AI in Software Engineering

Compared to our survey from last year, not much has fundamentally changed in terms of the AI tools landscape. However, we are seeing more interactive "rubber ducking" with AI, and teams are starting to experiment with AI agents that can take on more complex tasks. As these tools continue to mature, it's clear they will play an increasingly important role in software development.

But as with any new technology, it's crucial to approach AI with a critical eye and not blindly trust its output.We'll continue to monitor this space closely and keep you updated on the latest developments.

In the meantime, we'd love to hear about your own experiences with AI tools – feel free to reply to this email and share your thoughts!

Happy coding,

Arun & The NullPointerClub Team