IBM will use its new AI platform watsonx to bring mainframes into the modern age by translating Cobol into a more dynamic programming language.

Mainframes sound like dark age technology, but banks and government institutions still have these big computer boxes. However, critical applications running on a mainframe stand in the way of rapid digital transformation. IBM aims to solve this problem by unleashing the hottest technology on mainframes today: AI.

In May, IBM introduced watsonx, a new overall platform for artificial intelligence. One of the components of watsonx is Code Assistant, a tool that can generate code. With Code Assistant, companies can have critical mainframe applications rewritten in Java or any other modern programming language without leaving the computer room.

cobol

Mainframe applications are often written in Cobol, an ancient programming language that is rarely used to develop new applications today. For this reason, developers today no longer specialize in Cobol.

However, partly due to its widespread presence on mainframes, Cobol is still part of today’s IT landscape. Watsonx needs to bridge this generation gap and ensure that mainframe applications are translated into a language younger programmers can understand.

The hand that feeds you

“Never bite the hand that feeds you” is an old adage, and IBM certainly has no intention of making mainframes go away. Mainframes still make IBM a lot of money because it’s so difficult to port applications to another platform. Organizations also see benefits in the security of critical business data.

Nevertheless, there are also good arguments for the phasing out of mainframes. Mainframes don’t integrate at all with the rest of your IT ecosystem, and maintaining such a computer system can easily cost hundreds of thousands of dollars a year. The big cloud providers are trying to convince the last few mainframe users to pull the plug. For example, Google Cloud offers a “cut-and-paste” feature for copying data on mainframes to the cloud.