3 ms·

How would you apply that observation to Wikipedia? That's the business model we're aiming to take - crowdsourced, crowdfunded, non-profit, and community govern

How would you apply that observation to Wikipedia? That's the business model we're aiming to take - crowdsourced, crowdfunded, non-profit, and community governed. And the reason for that is precisely because of the mis-aligned incentives in both the pay to access and pay to publish models. The filtering here would be accomplished by the tagging system and the voting system.

jpmattia 4y ago
> How would you apply that observation to Wikipedia? I would notice that the Wikipedia reader base is a multiple of peer-reviewed scientific-paper base. While I don't know what the multiple is, I would guess the range is 1-2 orders of magnitude. Wikipedia claims its average donation is $15 [1], so a rough guess will imply $100-$1000 average donation for a sci pub model. That's on the income side. On the spending side: I would guess the major difference in costs would be the scientific editors, which exist in the traditional science publication model. There are probably large initial marketing costs associated with 1. finding the right target authors and convincing them to send a valuable contribution to the new site, 2. finding serious editors to manage peer review and separate the wheat from the chaff. (edit: I assume this is still part of the process. I think about something like Einstein's special relativity paper, and I'd have to guess it would have been widely downvoted as an internet crank trying to publish crazy ideas -- he did not have a reputation yet.) I see this as potentially expensive: The level of the first papers will determine how likely it is to succeed. Said differently: If the new site starts by publishing lame work, it's not like someone is going to want to publish there afterward. In any case: I think it is a tough and worthwhile problem to solve, and I wish you good luck. [1] https://donate.wikimedia.org/wiki/FAQ https://donate.wikimedia.org/wiki/FAQ
CJefferson 4y ago
The problem is wikipedia is explictly rejecting what you want to accept -- original research. It takes minutes to check someone's reference says what that it claims they say. It takes days to check if a new piece of independant research is correct.

jpmattia 4y ago

> How would you apply that observation to Wikipedia? I would notice that the Wikipedia reader base is a multiple of peer-reviewed scientific-paper base. While I don't know what the multiple is, I would guess the range is 1-2 orders of magnitude. Wikipedia claims its average donation is $15 [1], so a rough guess will imply $100-$1000 average donation for a sci pub model. That's on the income side. On the spending side: I would guess the major difference in costs would be the scientific editors, which exist in the traditional science publication model. There are probably large initial marketing costs associated with 1. finding the right target authors and convincing them to send a valuable contribution to the new site, 2. finding serious editors to manage peer review and separate the wheat from the chaff. (edit: I assume this is still part of the process. I think about something like Einstein's special relativity paper, and I'd have to guess it would have been widely downvoted as an internet crank trying to publish crazy ideas -- he did not have a reputation yet.) I see this as potentially expensive: The level of the first papers will determine how likely it is to succeed. Said differently: If the new site starts by publishing lame work, it's not like someone is going to want to publish there afterward. In any case: I think it is a tough and worthwhile problem to solve, and I wish you good luck. [1] https://donate.wikimedia.org/wiki/FAQ https://donate.wikimedia.org/wiki/FAQ

CJefferson 4y ago

The problem is wikipedia is explictly rejecting what you want to accept -- original research. It takes minutes to check someone's reference says what that it claims they say. It takes days to check if a new piece of independant research is correct.