3 ms·
I've given a lot of though about current state of journalism and my conclusion was that the current model is wrong for two reasons 1. conflict of interests 2.
by lukasm 13y ago
I've given a lot of though about current state of journalism and my conclusion was that the current model is wrong for two reasons
1. conflict of interests
2. lack of transparency
@1 There are two main sources of capital for journalists - government or companies that buy ads. The problem is when they can put pressure to not write of unjust acts that may conflict with their interests. The solution is to provide founding from large enough group of people when there is no dominant subgroup. Models has to be highly distributed to avoid bribing users[1]. There needs to be upper limit of funding e.g. 99$ or 10%[2] to dilute influence of a single user. Even better for a 5$ a month I can get credit which I can allocate.
@2 Log everything. Brutal transparency. That's the only way I can think of.
So you're going to make something people want, right? :) I want to be able to vote what is important for me. I want to post an issue and let other judge if it's worth investigating.
[1] I'm amazed how often pattern of distributed models occurs and works very well (Bitcoin, Git etc.).
[2] That may have a similar effect as Twitter's 140 chars limit
- SandersAK 13y agoYep, you've kinda nailed it. We're moving towards that on Beacon. But right now our primary focus is on helping writers convert their existing readership into paying subscribers. As the content offering as a whole becomes more robust, you'll start seeing more of the effects you're looking for. It's really simple and old fashioned: You think certain information or perspectives are valuable, you pay for them. We're just updating the underlying tech and user experience to modern times ;)