AI Isn’t the Hard Part. Governance Is. #platformengineering #ai

The problem with getting value with AI in enterprises, large organizations, is not about the models or even tools like cloud code or whatever else to get from it. It's about securing and governing it. It's about making it safe in the enterprise sense to use. Having controls over where things run, who accesses what, the cost that you incur, all these sorts of things that don't have anything to do with that core application or that workload that you're running. And the issue is that those are difficult problems to solve. And they often require new platforms, new technologies, and new ways of working. They require culture change, organization change. If you remember with DevOps and then platform engineering and S, we obsessed over all the trimmings, all that other stuff. So what you need to do if you want to kind of, if you'll pardon the phrase, unlock the value of AI in enterprises, is focus on how you make it secure and compliant to run. How you put the entire system in place that allows this thing that already works well and is kind of undeniable to just do its job.