This hit me as a key statement: "Without explicit leadership accountability for fraud outcomes—not just delivery milestones—the rational bureaucratic choice is to get the system running and treat the fraud controls as phase two."
And that's why we often get stuck in these doom loops.
A serious question for you. As someone who worked at Sabre, where we were migrating code from the mainframe to the cloud using AI, why not do the same with COBOL and Assembler (Assembly?) language? Am I simplifying this too much? I know it takes time. God knows Sabre is still doing it, but it can be done. Is it just the scale of this thing?
I think that’s being attempted now. But the core issue is that the data on these mainframes is so vitally important that losing it would be catastrophic. The stakes are so high no one is willing to risk losing it.
This hit me as a key statement: "Without explicit leadership accountability for fraud outcomes—not just delivery milestones—the rational bureaucratic choice is to get the system running and treat the fraud controls as phase two."
And that's why we often get stuck in these doom loops.
A serious question for you. As someone who worked at Sabre, where we were migrating code from the mainframe to the cloud using AI, why not do the same with COBOL and Assembler (Assembly?) language? Am I simplifying this too much? I know it takes time. God knows Sabre is still doing it, but it can be done. Is it just the scale of this thing?
I think that’s being attempted now. But the core issue is that the data on these mainframes is so vitally important that losing it would be catastrophic. The stakes are so high no one is willing to risk losing it.