While AI-suggested code still contains many bugs today, time-savings for developers lie in other things AI can help with, according to a new study.
Four academics from the University of Lugano in Switzerland discuss this Research paper Artificial Intelligence Transformation for Developers. They are focusing more specifically on their daily routine, and their workload is expected to halve by 2030.
In their paper they point to HyperAssistant, an intelligent evolution of GitHub Copilot, OpenAI ChatGPT, Google Gemini or Tabnine. To help, it focuses on the developer’s mental health by suggesting breaks or activities. Future AI will coordinate meetings with relevant team members and provide support in learning new skills.
“We envision HyperAssistant, an augmented AI tool that will provide comprehensive support to developers in 2030, addressing current limitations in mental health support, bug detection, code optimization, team interaction, and skill development,” the authors explain.
“We emphasize AI as a complementary force, augmenting the capabilities of developers rather than replacing them, creating advanced, reliable and secure software solutions.”
Chart of the situation in 2030
The paper cites an example of daily workflow in the future: “Ashley, a developer in the year 2030, comes into the office and immediately notices that some code has changed since yesterday. However, thanks to HyperAssistant, you’ll see a brief summary, highlighting only the relevant changes. Thanks to this efficiency, they quickly understand updates and are ready to start their tasks.
“When she starts programming, an intelligent error detection system alerts her to a mistake she has accidentally made. The system not only reports the error but also suggests possible solutions, simplifying the debugging process.
Read also
Stack Overflow sees a willingness to use AI, but no confidence yet
HyperAssistant goes so far as to schedule a meeting between Ashley and the senior developer, complete with relevant documentation. Along the way, the AI monitors for typos or other data that might indicate a need to take a break. With this help, the authors suggest that Ashley would only need half a day in 2030 to do what it would take her entire day in 2024.
It’s a wonderful drawing of the future. If you want to read the full study, you can here Justifiably. Now let’s hope like, Record He writes that it remains HyperAssistant and not HyperManager.
