Large language models excel at language tasks that benefit from statistical patterns. So from at the very beginning of GenAI, generating program code was an obvious application. As in many other fields, people stated that there was much more to programming than just writing code. Yes, and then came AI agents.
The speed of development of algorithms to generate code and managing all documentation and deployment processes is (again) impressive. After first milestones like GitHub Copilot, specialized IDEs like Cursor AI or Windsurf (formerly Codeium, acquired by OpenAI) showed what the future could look like. And now, with the introduction of Codex by OpenAI, the future seems here.
It’s no longer about generating code, it’s about but full-stack software engineering with integrated DevOps: GenAI has become a huge software factory that only needs few human operators to monitor (and align). And now? Instead of evaluation a standard software or a Saas platform to be ta8iilored to my needs, I rather quickly build it (or have it built). But not only the buying cycle integration work (any SAP consultants reading?) will change, but also the lifecycle management of software.
And yes, some people will need silver spoons that will last for generations, but many people like fast food and do eat it using cutlery made from cheap (and polluting) single-use plastics. Welcome to the world of tactically build individual software.
