Leverage Artificial Intelligence to Significantly Increase Software Developer Productivity

To become productive and competent enough, developers new to software applications often require months, even years, of on-the-job training to avoid making dangerous mistakes that put systems at risk. Discerning the true intent of the functionality in the code is no easy task, and today’s developers spend approximately 75% of their time searching the source code to identify the code that represents the functionality that needs to be changed or modernized.

While code search tools, linters, and static and dynamic analysis tools can help developers significantly improve their efficiency and effectiveness, many of these tools fall short when it comes to actually identifying the specific lines of code that require attention, especially with regard to how tangled throughout a system that code often is.

Current approaches employed by today’s AI are helping to improve this challenge, but they fall short in many important ways. Fortunately, employing a novel approach to AI vastly improves a developer’s ability to efficiently identify specific lines of code relevant to tasks that they spend an enormous amount of time on every day.

The shortcomings of current tools

As software repositories continue to grow uncontrollably to unprecedented sizes, those responsible for maintaining and preserving the functionality of a system say that it is increasingly difficult to find errors in the code without the assistance of the machine. Even worse, debugging takes about half a developer’s time, and correcting and then validating the correction of a single error can take days, weeks, or even longer.

From troubleshooting to program repair to code synthesis, today’s tools can be helpful in analyzing millions of lines of code, flagging bugs, and suggesting fixes or best practices, but a number of search tool challenges remain. existing code. Code navigation in a tool like Sourcegraph accelerates the rate at which developers can build a mental model of the code, helping them understand the parts of the code universe that are most relevant and important, but the tool doesn’t actually build the model. mental. . The tool only brings the developer closer to code that might be relevant to the intended behavior, but the developer must rebuild that mental model. As painstaking, mentally challenging, and full of mistakes that effort can be, human beings are ultimately fallible even if they are using the best tools available to help them.

For example, security vulnerabilities are simple enough to identify with current tools, but focusing on all the relevant code to fix is ​​difficult and time-consuming, much more than writing the program in the first place. Often the relevant code is scattered across many files, adding to the complexity of the task. When debugging unknown programs, developers often start the process with a “why” question about the behavior of their program, but have to conceptualize that question in a series of actions and queries to address the problem at hand. As a result, programmers must simultaneously track multiple aspects of the detailed program execution path; however, the ability to do so is incredibly challenging for even the most seasoned developers.

If a programmer does not understand how changing the code in one area of ​​the code affects the system as a whole, even a small adjustment can break the entire system. Since it is important for developers to understand what an application actually does, debuggers are not as widely used as you might expect, especially in functional programming environments.

Unfortunately, even the most advanced static and dynamic analysis tools, search tools, and linters are extremely inefficient at conveying that knowledge. Whether it’s drawing pretty pictures, tracing code, seeing how code behaves with specific data as input, or providing a summary of searched words followed by complex changes, these tools help developers identify relevant code in different ways, but the developers have yet to do so. conceptualize to expose the intent of previous developers encoded in the software and reveal the bug they seek to correct. At the end of the day, even with the most advanced tools, human software developers still have to build a mental model of what the code does by deciphering the code directly to match the behavior they seek to change so that they can effectively modify it. complex and critical systems.

Harnessing the power of AI to overcome shortcomings

From Intel to Amazon to Microsoft, some of the world’s biggest tech companies are turning to and creating AI-powered tools that propose snippets of code to developers as they code. By leveraging artificial intelligence, developers are witnessing productivity gains, a reduction in time spent writing code, and risk mitigation.

However, by using AI in a novel way, the code repository can become a repository of knowledge that represents source code in the same way that a human being thinks about the world, in cause and effect. When AI-powered tools know each and every behavior throughout the source code, the tools can “collaborate” with developers to limit the code that needs to change, without looking for and understanding all the surrounding code. Using AI in this way means increasing developer productivity by orders of magnitude.

By allowing developers to request specific code for a behavior from this knowledge repository, AI helps developers deal with only the relevant code, allowing them to make changes with the confidence that they won’t break the entire system. Even better, developers can “propose” a change while they are coding without recompiling or verifying the code, so the use of AI mitigates the risk by verifying if the change is isolated. This ability also extends to the CI / CD pipeline by having the AI ​​verify the code after each build to ensure that the behavior will never change in the future.

Employing AI in this way focuses the developer on only the code that matters, without the undue cognitive burden of understanding code that is irrelevant to the task at hand. Unlike code search tools, linters, and static and dynamic analysis tools, AI ultimately enables developers to make necessary changes safely, efficiently, and effectively..

Steve Brothers

Steve Brothers

Steve joined Phase Change as COO in 2018, bringing more than 30 years of experience in technology-related organizations with sales, technical and leadership roles in industries such as financial services, healthcare and services. Previously, Steve held positions as CEO of Ajubeo and Executive Vice President and CIO of Urban Lending Solutions. Steve graduated from the University of Colorado at Boulder and has a BA in Philosophy and a BA in Information Systems. Steve is a proud father of two, is a mentor at Galvanize and resides in Golden, CO.


Leave a Reply

Your email address will not be published. Required fields are marked *