Amazon CEO Andy Jassy says the company's AI assistant Amazon Q has dramatically cut the time needed for Java updates. Amazon estimates this saved thousands of developer years and hundreds of millions of dollars.
In a LinkedIn post, Jassy highlighted efficiency gains from using Amazon Q for software development. The AI tool significantly reduced time and costs for updating Java applications to newer versions.
Jassy called updating core software one of the most tedious but crucial tasks for development teams. Since it doesn't add new features or visibly improve user experience, teams often avoid or delay these updates in favor of new projects.
Using a new code transformation feature, Amazon Q cut the average time to upgrade an app to Java 17 from about 50 developer days to just a few hours, Jassy said. The company estimates this saved about 4,500 developer years.
Amazon Q Code Transformation analyzes existing code, suggests changes, and implements them. It updates package dependencies, revises outdated and inefficient code, and integrates security practices.
Nearly 80 percent of AI-generated code used without changes
In six months, Amazon updated over half its Java production systems to newer versions much faster and with less effort. Developers used 79 percent of Amazon Q's auto-generated code reviews without changes.
Jassy said the benefits go beyond saved development time: The updates improved security and lowered infrastructure costs, leading to estimated yearly efficiency gains of $260 million.
He sees this as proof that large companies can achieve major efficiency gains in maintaining core software by using AI. For Amazon, it was a "game changer."
Of course, keep in mind that Jassy is trying to sell his company's AI software here. But even if Amazon's estimates are way off and the actual savings are a fraction of what Jassy claims, the numbers would still be significant.
Then again, others have criticized AI code for being crap and creating more work than it solves. We won't know until some neutral long-term studies of AI's impact on coding come out.
