There are many examples where decentralization (open architecture) has destroyed established industries in a relatively short period of time. For example, Linux vs. other operating systems. People often don't realize that 96% of the Internet is powered by Linux, their car, refrigerator and alarm clock all run on Linux. Supercomputers, AI are all 100% Linux. It only took 10 years for Linux to crash companies with tens of thousands of developers.
Another example: Internet versus Intranet. The Internet was portrayed as an unreliable, hacker-infested dumpster, while (expensive) secure Intranets were there to save the world. Where are they today? When was the last time a document was updated on one of these?
What about enterprise software? Remember when system administrators used to configure all employees' computers and install advanced software? Today, most innovations are done on employees' smartphones, where users install the applications themselves, and system administrators are trying to catch up with recent trends.
That is why I am optimistic about decentralized and open Hive social blockchain taking on the established social networks.
Hive is not truly decentralized. It is a semi decentralized but that's necessary for the scalability requirements.
Decentralizations means you don't need to ask one's permission to build apps, communities and other projects on the chain. Anyone can build apps on Hive, use all the user data, entire social graph, objects, attention bids available on the chain. Permissionless is the fundamental requirement for rapid innovation.
In terms of distribution of HIVE ownership, it is also rather decentralized. 30% of HIVE is now owned by 35 accounts. In the world of blockchains, this is very much decentralized.
Lol, it only takes 15 people to change that in Hive. That's why I am calling it semi decentralized.
Ok, so you are talking about the DPOS itself and the 20 top witnesses (and 15 of them can implement a hard fork). Well, this is a long discussion with lots of good arguments on both sides.
But what I want to focus on, is that true decentralization in that respect (implementing code updates) comes at a huge cost to innovation. We are nowhere close (blockchain base code) to the final version. We are still heavily experimenting. We solved many attack vectors, but we still need SMTs and a bunch of other things.
Once you'll have a blockchain version with a consensus that you would like to freeze for longer, sure, extend the list of witnesses to 200, reduce number of simultaneous witness votes and your are done. It will be super hard to introduce the next upgrade (change rules) after that.