I heard this narrative before. Three thoughts: - People concentrate on the tool (technology) vs. goal (value) Let say you need to handle a million users in th

I heard this narrative before. Three thoughts: - People concentrate on the tool (technology) vs. goal (value) Let say you need to handle a million users in the past (read: "provide value to million users"). It would be an unbelievably huge effort back in 1996, requiring _A LOT_ of logistics. For most companies, it would take years to reach out to these million users. And god forbid you to need to send out an update. Fast track to 2021. One person can put the code in a docker, drop it in AWS Kubernetes, scale it up and be able to serve million users by the end of the day. (BTW. I am not saying that you don't need to do marketing/sales and so on, but purely engineering/distribution end up being way simpler). Oh... BTW. All these people can use this new service on the go, because everybody has a computer in his pocket with access to the internet. And get update to it pretty much instantaniously. I don't care that a new language has curly brackets or a new programming paradigm invented each year or once a decade. I care how much value could be created. And tools/languages/frameworks/infrastructure in 2021 provides a way to build way-way more value - You don't need to improve a hammer that much The first hammers date back ~3 millions year ago. There were relatively few improvements to it for a looong time. The idea is still the same (a heavy head and a long arm). If the tool works well, you don't need to reinvent it that much. You invent other, more complex, and specialized tools. - Ignoring real achievements There were dozens of things mentioned here in this thread, which wasn't mentioned in OP (LLVM, Mobile, ML, and dozens and dozens of others).