Intel’s principal AI researcher, recently published an article on LinkedIn with the news of the public release of ControlFlag, their machine learning-based software that can detect problems in computer code. The researchers assert it has found hundreds of defects and vulnerabilities throughout production quality apps with no human assistance required.
Intel Labs are using Machine-learning techniques in an effort to reduce the time needed for debugging. The ControlFlag uses self-supervised machine learning that autonomously detects coding anomalies, cutting down on systematic human review and improving quality at scale.
ControlFlag has been extensively tested and is now widely used in production-level software. For example, last year, it identified an error with Client URL (cURL), a computer program transferring data using various network protocols over one billion times per day – this was reported to the cURL team, which then patched their code as agreed upon by Control Flag’s findings.
In a world where almost all software has bugs, machine programming (MP) is the solution. Machine-programmed AI algorithms can find even those human programmers might have missed and keep your programs running smoothly without any hiccups or crashes.
ControlFlag is a new AI that can work with any programming language and utilizes the emerging concept of semi-trust to learn from unlabeled source code. As more data becomes available, it evolves into what makes itself better by its own accord.
The ControlFlag project has the potential to save countless hours of developers’ time and money. This, in turn, will allow them – as an entire community-to accelerate technology advancement even further.
Intel Article: https://www.intel.com/content/www/us/en/newsroom/news/machine-programming-tool-detects-bugs-code.html