> With AWS, there is no infra to manage like in companies of yesteryear; it's all abstracted away to AWS. I would argue that this is an extraordinary simplific

> With AWS, there is no infra to manage like in companies of yesteryear; it's all abstracted away to AWS. I would argue that this is an extraordinary simplification. Infrastructure is abstracted away to the cloud vendor insofar as you don’t (typically) need to worry about power, physical connectivity to the host, logical networking at the layer of (for example) BGP, backup power, and so on. But you absolutely have to concern yourself with infrastructure management to some extent. Be it backups, CI/CD infrastructure, BCDR management, geographic redundancy, proprietary vendor software that has prerequisites that your cloud provider doesn’t manage, OS patching for VMs, and so on. Even if a business is 100% on a serverless stack for its service, it will still find itself with infrastructure to manage somewhere. Some businesses have infrastructure because they need some type of interoperability to failsafe between clouds. Even if a software business could succeed with no infrastructure of its own and no system administrators, that doesn’t have anything to do with tech debt. Tech debt can and does happen in any stack.