I was watching the Microsoft Ignite event on their event site earlier. It was interesting but not as impactful to me as last year’s Ignite. But there was this thing that Scott Guthrie said something in Microsoft Cloud in the era of AI which grabbed me:
GitHub Copilot dramatically accelerates developer productivity. 46% of all lines of code written by developers using GitHub Copilot are now fully AI-generated. And 96% of developers report being able to complete repetitive tasks much faster. And 75% of developers say that GitHub Copilot helps them focus on more satisfying work and enjoy their jobs more. I love this last statistic that GitHub Copilot doesn't just bring productivity, it also brings meaning and satisfaction the most. That's the real power and potential of what AI provides.
A few points here: One is the speed of adoption is wild. But to know that implicitly means you have to also know LLMs have only been out almost exactly a year now. But that 46% of all lines of code written by developers using GitHub Copilot are now fully AI-generated is just crazy. I’ve messed around with ChatGPT, and a lot of that code is… not great. It’s fast – way faster than I could but what it spits out is often barely above spaghetti code.
I have no doubt it can complete a repetitive task quickly, but there are other ways of doing that with less risk. As for the 75% stat about satisfaction, I think those are two different points. You can be focused intently on something which is ultimately satisfying. That’s a hallmark of professionalism. Also, much of job enjoyment is related to things like team dynamics or company culture.
The bigger point is, in spite of the metrics, this is all subjective. They’re not saying people are completing tasks any percentage quicker or with fewer errors. They’re not even using lines of code generated which is a pretty bad metric anyway. They’re just presenting how people feel about it.
And that makes me wonder.