Using AI tools to debug computer code is becoming increasingly common among programmers at large companies, according to a recent survey conducted by Microsoft's GitHub. Out of 500 programmers surveyed, an overwhelming 92% reported utilizing AI tools in their workflow. Wakefield Research, a PR polling firm based in Virginia, was hired by GitHub to conduct the survey and analyze the results. The researchers discovered that 70% of respondents recognized the benefits of using these tools.
While AI coding tools can help developers create and debug code more efficiently, the survey revealed that measuring productivity solely based on code volume may not be the most accurate metric. GitHub noted that developers praised AI tools for assisting them in meeting performance standards by improving code quality, accelerating outputs, and reducing production-level incidents. As AI tooling becomes more prevalent in software development, engineering leaders must reconsider whether code volume is still the most appropriate measure of productivity and output.
Stack Overflow, a popular Q&A forum for programmers, has also experienced a decline in traffic due to the increasing reliance on AI. According to web analytics firm SimilarWeb, Stack Overflow's traffic has steadily decreased since the beginning of 2022, with an average monthly drop of 6%. In March, the decline was particularly notable, with a 13.9% decrease compared to February. Similarly, April saw a 17.7% drop from the previous month. Programmers are turning to ChatGPT, the viral chatbot developed by OpenAI, for immediate solutions to their coding queries. However, it's worth noting that ChatGPT's responses often stem from data scraped from Stack Overflow.
While the surveyed developers primarily see AI tools as a supplement to their current work, the possibility of AI replacing human employees looms on the horizon. A recent workforce report revealed that nearly 4,000 jobs were lost in May due to the implementation of artificial intelligence. With the prospect of an impending recession, companies may be inclined to downsize their human workforce in favor of cost-saving AI solutions to appease shareholders. Nevertheless, these companies should acknowledge that AI has the potential to enhance labor by streamlining menial tasks or even eliminating them altogether, thereby optimizing the performance of both AI and human workers. Research from Stanford University and the Massachusetts Institute of Technology indicates that 14% of employees incorporating ChatGPT into their workflow experienced increased productivity, with the least experienced and skilled workers completing tasks 35% faster.
