Surveys on the impact of AI on software development over the past few years (AI column)
Surveys on the impact of AI on software development over the last few years (Rubric #AI)
Surveys are often the standard way to assess the state of an industry and contain valuable data. It is good when they are based on the answers of a large number of respondents. And since I am interested in the impact of AI on development, I decided to study what surveys have already been conducted, what results of their conduct, and what methodologies were used by the authors. In future posts, I will separately analyze interesting points, but for now I will share with you the resulting selection.
1. Stack Overflow Developer Surveys (2023, 2024) These surveys are among the largest and contain valuable data. They are based on a survey of tens of thousands of engineers still using Stack Overflow. (Not switching to ChatGPT completely.). Polls 2023 and 2024 Over the years, a separate block of questions about AI has been asked. I wonder what in 2023 year 70percentage 90,000 Developers have already used or planned to use AI tools in their work. K 2024 This figure has risen to 76%. The main tools became ChatGPT (83%) GitHub Copilot (56%). These surveys can be used to track trends, as the questions are similar from year to year. By the way, polling. 2025 It has already passed, but the results have not yet been published. When I get there, I'll spin the data and look at the trends.
2. GitHub Developer Survey (August 2024) GitHub conducted a study among 2,000 Developers from the USA, Brazil, India and Germany. The results showed that 97The percentage of respondents used AI tools at work. Notably, 61-73Percentage of developers, depending on the country, believe that AI enhances their ability to create software. There. earlier investigation of GitHub September 2022 Its results are interesting for the history and study of methodology, but the information itself is not relevant to the current situation.
3. JetBrains Developer Ecosystem Survey (2023, 2024) JetBrains is conducting research with 2017 For the first time in the year, we included questions about AI in 2023 year. In 2024 year 84% of developers were familiar with generative AI tools. The study found that developers are most interested in delegating AI routine tasks.
4. DORA Research (2023, 2024) Dora Metrics Created by DevOps Research and Assessment Team (DORA)This is the case with Google right now. Each year they publish reports and 2025 year published Impact of Gen AI in Software Development. They found that 76Percentage of developers use AI for daily tasks The paradox is that AI increases individual productivity. 2.1%, but reduces team performance by 1.5%. I will analyze this report separately in more detail.
5. WeAreDevelopers "State of WebDev AI" (2025) Interesting. studyIn which 4K+ respondents participated. 91Percentage of web developers surveyed “experimented” with ChatGPT; Cursor IDE recognized as the leader of AI editors (80% recognition) .
6. The Pragmatic Engineer Survey (2025) Interesting. studyIn which 3K+ respondents participated. 72% of engineers consider AI tuning to be the “most innovative market”; most note the rapid arrival of startup tools competing with Copilot
Research shows that AI has become an integral part of software development, but the industry is still in the process of adapting to this technology, balancing benefits and risks.
P.S. I recently made a selection of this. (1 and 2) topically **"Developer Productivity for Humans"**Each of which I studied and studied separately. I think that the surveys from today’s collection I will analyze in detail later:)
#AI #ML #Software #Engineering #Processes #Metrics #Management