In a groundbreaking study, researchers at the University of Stirling have harnessed the power of artificial intelligence (AI) to create faster and more reliable software. Dr. Sandy Brownlee, leading the study, employed ChatGPT, a state-of-the-art large language model, to automatically optimize computer code, resulting in significant improvements in software performance.
Revolutionizing software development
Software development has long been a challenging and time-consuming endeavor. The need for faster and more reliable software has become increasingly pressing, especially as the complexity of applications continues to grow. Dr. Brownlee’s team at the University of Stirling set out to revolutionize the field by leveraging the capabilities of AI, specifically ChatGPT, to streamline and enhance software development.
ChatGPT software optimization
The research led by Dr. Brownlee employed ChatGPT to automatically analyze and optimize computer code. By requesting the AI model to suggest improvements, the researchers aimed to create software that runs faster and more reliably. The results of this innovative approach have the potential to impact a wide range of applications, including mobile apps.
Dr. Brownlee explained,
“We found that, on the open-source project we used as a case study, a large language model was able to produce faster versions of the program around 15% of the time, which is half as good again as the previous approach.”
One of the most remarkable aspects of the study was ChatGPT’s ability to identify and improve areas of the software that had not been explicitly highlighted by the researchers. This adaptability and creativity in optimizing code make AI-driven enhancements even more valuable.
Dr. Brownlee noted also the most tangible benefit is in your pocket – mobile apps that run more efficiently mean that your battery lasts longer, and the apps will be more responsive when in use
In addition to improving user experience, this research holds significant implications for sustainability. As software complexity increases, computers consume more energy, contributing to environmental concerns. The ability to efficiently automate software improvements could play a crucial role in helping countries achieve their net-zero commitments.
Dr. Brownlee emphasized the potential trade-off between the energy required to develop and operate large language models and the energy savings they can generate when applied to widely used software. He stated,
“There is a trade-off here because large language models cost a lot of energy to make and use, but if they can be used to improve a piece of software that is run many times, the payoff may be worth it.”
Transforming software development
The research led by Dr. Brownlee has the potential to transform the landscape of software development. With the advent of automated software improvement powered by AI, the roles of software developers may evolve, allowing them to focus on higher-level design tasks. This trend aligns with the direction software development has been heading for decades.
He also noted that it’s quite possible to be concerned about all this – AI writing computer code sounds like the beginning of a sci-fi disaster plot – but this is not going to lead to an AI takeover,
He reassured making tools that support developers by providing more information, and it’s still critical to have a human involved to use that information effectively.”
The research findings, titled “Enhancing Genetic Improvement Mutations Using Large Language Models,” were presented by Dr. Brownlee at the Symposium on Search-Based Software Engineering 2023 in San Francisco, part of the ACM Joint European Software Engineering Conference and Symposium on the Foundations of Software Engineering.
The study was a collaborative effort, involving researchers from various institutions, including University College London, Johannes Gutenberg University, and King’s College London. The project received support from a UKRI EPSRC grant and the ERC advanced fellowship grant.
The groundbreaking research conducted by Dr. Sandy Brownlee and his team at the University of Stirling has the potential to usher in a new era of software development, making software faster, more efficient, and environmentally friendly. As AI continues to play an increasingly prominent role in various fields, its application in software optimization promises to shape the future of technology for the better.
