3 ms·

Take the 5 billion, give it to a decentralised opensource FB clone project. Same about 100 million for a global advertising campaign.

by leke 8y ago

djohnston 8y ago
social networks with a lot more capital have tried and failed. it's the network effect and the inertia of migrating that keep people on the network, not the lack of a "fbclone"
- AnthonyMouse 8y ago
  The other social networks with that amount of capital were trying to be Facebook, i.e. another closed network that they can profit from. Facebook just under a different name has no relative user advantage and no reason for users to switch. An actual decentralized network has different requirements and can thereby attract users by having different characteristic advantages. For example, one of the large and growing problems with centralized networks is that everybody leans on them to do censorship, and they have insufficient stake in protecting individual speakers, even against unreasonable or blatantly fraudulent censorship requests. If you make it trivially easy and inexpensive for anyone to self-host then the person making the censorship decision is the person whose content it is to begin with, which brings back the proper alignment of incentives regarding cost of removing content vs. liability for hosting it. And that's just one advantage. Another is you control your data instead of Facebook and it's only you and your own friends who have access to it, not some corporation using it to run adversarial machine learning against you. Another is that your data is local and is therefore faster and available in places with poor internet. Another is that you have access to it yourself and can make modifications to both code and data, and then you get improvements from everyone instead of being restricted to a single company with a conflict of interest. These kinds of things add up, but first you need somebody to spend the resources to develop it into something that non-experts can easily use and understand. Which is already happening, slowly, but would happen much faster with those kinds of resources.

Take the 5 billion, give it to a decentralised opensource FB clone project. Same about 100 million for a global advertising campaign.

by leke 8y ago

Take the 5 billion, give it to a decentralised opensource FB clone project. Same about 100 million for a global advertising campaign.

djohnston 8y ago

social networks with a lot more capital have tried and failed. it's the network effect and the inertia of migrating that keep people on the network, not the lack of a "fbclone"

AnthonyMouse 8y ago
The other social networks with that amount of capital were trying to be Facebook, i.e. another closed network that they can profit from. Facebook just under a different name has no relative user advantage and no reason for users to switch. An actual decentralized network has different requirements and can thereby attract users by having different characteristic advantages. For example, one of the large and growing problems with centralized networks is that everybody leans on them to do censorship, and they have insufficient stake in protecting individual speakers, even against unreasonable or blatantly fraudulent censorship requests. If you make it trivially easy and inexpensive for anyone to self-host then the person making the censorship decision is the person whose content it is to begin with, which brings back the proper alignment of incentives regarding cost of removing content vs. liability for hosting it. And that's just one advantage. Another is you control your data instead of Facebook and it's only you and your own friends who have access to it, not some corporation using it to run adversarial machine learning against you. Another is that your data is local and is therefore faster and available in places with poor internet. Another is that you have access to it yourself and can make modifications to both code and data, and then you get improvements from everyone instead of being restricted to a single company with a conflict of interest. These kinds of things add up, but first you need somebody to spend the resources to develop it into something that non-experts can easily use and understand. Which is already happening, slowly, but would happen much faster with those kinds of resources.