6 thoughts on “Introducing Decentralized Autonomous Wikis”
Interesting. Questions (I’m no expert)…
Could the history and all changes of a wiki be regarded as a blockchain?
Yes, probably.
Could editing and rating wiki pages serve as mining?
No. Mining must be comptuationally (and therefore financially) expensive to ensure that real work was done. Plus, there has to be a method of increasing difficulty as time goes on.
How to keep out attackers?
This relates to the point above. Attackers are intrinsically kept out if there is real financial cost to participate. There would be an interesting additional market dynamic in this model, link juice. Spamming would have some real financial value if it would generate clicks and therefore eyeballs and therefore advertising dollars. Probably need some thought…
How to make sure that articles do not collide while at the same time allowing different “wiki realities”?
What is article collision?
What will be the incentive for mining?
I think this could also work if it’s merely based on the love of knowledge and cooperation. People have added 10M+ articles to Wikipedia without any monetary incentive.
Decentralized P2P systems can exist without mining. What about autonomy? Is Bittorrent autonomous? Is the internet itself autonomous like a DAC?
But then again, I just asked on #datacoin if there’s a testnet. I think hacking together something that can “upload” a wiki onto test-datacoin could be a good start at this point.
Storing 1 kB of data costs 0.05 DTC, less than 0.02$ now.
Considering this it could be good to charge cost price for link-less text and a lot for adding a link (say 10 DTC). If a link is obvious spam the link will be removed. If the link adds value half the money can be refunded. If text is accepted part of the link coins can go towards the text writer, to create an incentive. Of course this can all get quite complicated.
It also touches Jaron Lanier’s ideas, that people should get paid for data.
Could the history and all changes of a wiki be regarded as a blockchain?
Yes. It would simply be a record of changes.
Could editing and rating wiki pages serve as mining?
No. Hash values are judged objectively. Edits and ratings are judged subjectively. What stops someone from automatically rating and editing pages? Take rating, for example. You could easily build a bot that would automatically rate new edits. Good ratings should be rewarded (and probably bad ratings punished). The only way I see of doing that is by rewarding the ratings that vote on the winning side of an decision. If the the current number of votes is public, this is easier to automate.
Your challenge is this: to build a decentralised way of rewarding efforts that are judged subjectively.
How to keep out attackers?
I think you should try and use the proof-of-state model. Users who have more successful edits can have more powerful votes (although this might put the power back into the hands of a few wiki geeks).
Interesting. Questions (I’m no expert)…
Could the history and all changes of a wiki be regarded as a blockchain?
Yes, probably.
Could editing and rating wiki pages serve as mining?
No. Mining must be comptuationally (and therefore financially) expensive to ensure that real work was done. Plus, there has to be a method of increasing difficulty as time goes on.
How to keep out attackers?
This relates to the point above. Attackers are intrinsically kept out if there is real financial cost to participate. There would be an interesting additional market dynamic in this model, link juice. Spamming would have some real financial value if it would generate clicks and therefore eyeballs and therefore advertising dollars. Probably need some thought…
How to make sure that articles do not collide while at the same time allowing different “wiki realities”?
What is article collision?
What will be the incentive for mining?
I think this could also work if it’s merely based on the love of knowledge and cooperation. People have added 10M+ articles to Wikipedia without any monetary incentive.
Decentralized P2P systems can exist without mining. What about autonomy? Is Bittorrent autonomous? Is the internet itself autonomous like a DAC?
But then again, I just asked on #datacoin if there’s a testnet. I think hacking together something that can “upload” a wiki onto test-datacoin could be a good start at this point.
Storing 1 kB of data costs 0.05 DTC, less than 0.02$ now.
Considering this it could be good to charge cost price for link-less text and a lot for adding a link (say 10 DTC). If a link is obvious spam the link will be removed. If the link adds value half the money can be refunded. If text is accepted part of the link coins can go towards the text writer, to create an incentive. Of course this can all get quite complicated.
It also touches Jaron Lanier’s ideas, that people should get paid for data.
Could the history and all changes of a wiki be regarded as a blockchain?
Yes. It would simply be a record of changes.
Could editing and rating wiki pages serve as mining?
No. Hash values are judged objectively. Edits and ratings are judged subjectively. What stops someone from automatically rating and editing pages? Take rating, for example. You could easily build a bot that would automatically rate new edits. Good ratings should be rewarded (and probably bad ratings punished). The only way I see of doing that is by rewarding the ratings that vote on the winning side of an decision. If the the current number of votes is public, this is easier to automate.
Your challenge is this: to build a decentralised way of rewarding efforts that are judged subjectively.
How to keep out attackers?
I think you should try and use the proof-of-state model. Users who have more successful edits can have more powerful votes (although this might put the power back into the hands of a few wiki geeks).
Just bumped into http://twister.net.co/ – definitely worth looking into.