Nathen Harvey discusses a developer survey on AI tools: most report higher productivity and better code quality, but 30% don't trust AI-generated output. The takeaway is verify what AI produces, don't just accept it.
How do you feel as a developer or as really any role as a participant in our survey, which might be a developer, a product owner, a tech lead, a CTO, but how do you feel AI is impacting your productivity? And what you can see here is that a large majority of our respondents feel like it is impacting their productivity in a positive way. It's one thing for me to feel more productive, but how is it going with code, the code quality? But on the whole, we're seeing that folks feel like it's improving their code quality.
I think one of the number one questions when we're talking to customers about AI in the field is like, how does this actually impact our code quality? And so at least from the customer perspective, it appears that you're probably building things faster with AI, but you're not dealing with nearly as many errors.
So was my productivity better? Yes. And it definitely made the code quality better in the moment. We see from individuals that they feel more productive on the whole, that it's improving their code quality on the whole. About 30% of our survey respondents trust the output a little or not at all. I think that like with all software changes that we have, we should trust but verify, right?
