This AI predicts coding errors before programmers commit them

The complexity of the work of computer programmers and developers makes it more than essential to have a resource. That allows them to identify errors, which would mean an increase in their performance in all types of projects, especially about the time factor.

This AI predicts coding errors before programmers commit them, This AI predicts coding errors before programmers commit them, Optocrypto

This AI predicts coding errors before programmers commit them

That is what the French-based video game company Ubisoft wants to achieve through the creation of an Artificial Intelligence (AI). That, not only notifying about coding errors but also predicts them before they are committed.

Commit Assistant; this tool aims to predict errors before the programmers include them in the software of an individual game. It seems that Ubisoft has a database that provides code developed for ten years. That is a whole software library. Well, the AI created by the company, was able to analyze where the errors were committed in the code, as well as the corrections applied to them.

This Machine Learning technology  Commit Assistant was able to allow to make predictions about when a developer would be at risk of including a similar failure. Likewise, the company has prioritized the creation of algorithms of this type to improve its production. Especially since the elimination of errors in games is one of the most challenging milestones to achieve within this sector.

Ubisoft has declared that the fact of predicting failure before the game is officially launched is a challenge. And therefore, Commit Assistant is a tool that seeks to be infallible, being able to detect so far 6 out of 10 errors with total precision.

Now, although this solution is applied to the video game sector. But, the truth is that this tool could inspire the creation of other devices capable of detecting failures. That is during the software development process. The Commit Assistant project was created by an association between the University of Concordia, which will be responsible for the publication of academic articles about this technology. And serving as an inspiration to create similar projects.