I started Wikipedia back in 2001 and after a long string of nonprofits have started Infobitt.com, which is, sort of, Wikipedia for news. We scour the web for facts about a given story, aggregate hand-written summaries of the facts, rank-order them, and rank-order whole stories (collections of facts = bitts) as well. At scale, we will make news less noisy and more efficient to catch up on. My claim is that only a giant online community could do this--traditional news orgs aren't big enough, and algorithms aren't sophisticated enough (any more than they're sophisticated enough to write Wikipedia).
What about Wikinews? It is a Wikimedia project and does collaborative community written articles. http://www.wikinews.org/
It is bothering me that I have to even sign up to view anything on Infobitt.com. That already is sending me away since I refuse to blindly sign up for any service.
Actually I prefer it to many news formats. If there is a big story (e.g. https://en.wikipedia.org/wiki/2011_England_riots ) and you turn on the news or go to the front page of reuters.com in the middle of it they are usually focused on some up-to-the-second minutiae or the latest sub-story to have occurred in the larger story.
This is fine if you have been paying attention to everything up until that point.
This is not so great if you came in a bit late and want an overview of everything about the story.
We handle this problem by creating the page as a news stream, based on each individual news event. We don't have UK riots yet, but here's an equivalent. http://newslines.org/michael-brown/ Unlike Wikipedia's static page-based format, because each news item is discrete, readers can sort the line from oldest news to first, or vice versa. We also use video and other embedded media to tell the story of events, unlike Wikipedia's textbook-like pages.
I actually agree. This is one area in which Wikipedia shines. But when it comes to breaking news, or tracking the actual order of events as they were reported, not so much.
I have of course wondered what it would be like to do Infobitt with a wiki, and I considered setting a wiki up for that purpose. The bottom line is that wikis lack the potential reasons for using the Infobitt format in the first place--making it easier to compete as well as collaborate, making it possible to vote on small pieces of content (as well as the ordering of the content), etc.
In other words, wiki forces users to collaborate on the same extended piece of content. This has all sorts of great effects, if enough people are participating. But it makes it harder to make short fungible pieces of content, rearrange them by vote, and do contests to discover the best version of each type.
How do you aim to tackle news about Russia & Ukraine?
How do you aim to tackle lengthy bike-shedding over trivial nonsense? (for example, there are at least a million words about hyphens, dashes, endashes and em dashes).
It's democratic, like Hacker News. We'll ask people not to include value judgments/obvious bias, but when it comes to relative ranking of facts, and how facts are worded (a matter for contests), we'll have to trust in the moderating influence of a wide variety of contributors.
In the fullness of time, at scale, we'll have features permitting us to filter/weight the news presented to us according to our preferences. Breaking news into small pieces makes this possible.
Interesting. How do you plan to keep coverage unbiased? A lot of bias manages to sneak into Wikipedia articles on current events in spite of their community guidelines and moderation.
The fact that bitt and fact order, as well as which fact is displayed to the reader, is (or will soon be) subject to vote will provide a moderating influence, I hope.
The reason Wikipedia articles are often biased is that people in authority in the community simply do not respect the neutrality policy. Even if most Wikipedians and most readers would prefer things to be more balanced, those in de facto control of an article determine the article's degree of bias. If the wording and order of facts is determined by open public vote, my hypothesis is that there will be more of a moderating effect.
However it would depend on who is doing the voting. If you, for example, used Twitter to gauge political sentiment, you'd be representing x% of the majority position, which would introduce a selection bias that would imply that x% represent the views of everyone. The problems with Wikipedia for controversial topics Ali's that typically both sides are highly motivated to "win" whereas those that aren't so motivated probably have the correct answer.
I second this one. Traditional news have little problem with their reputation. In general it's known if the newspaper/magazine describes itself as leftleaning, conservative etc..
I can imagine this to be much harder in a community with a vague, biased crowd not identifiable due to the hight amount of uncoordinated accounts participating.
I'd love to hear more thoughts on this bias-problem.
Battling organized partisanship is a problem for down the road. My hope is that, by the time we deal with that, we'll have the funding and the personnel to code up a system that enables us to test out some technical solutions to this problem. There are lots of ideas...
It's not only feasible, we are already doing it! [1] I have also written extensively on why Wikipedia is not a newspaper [2] and the many flaws of Wikipedia for news-based content [2]
Larry's attempt to break down the news into bits is an interesting experiment (and I wish him the best of luck as he goes forward). I'm not sure he will be able to attract the user base he needs to make it work though. We pay our writers, and are moving to implementing a revenue-share model to reward our writers. I'm not sure if the "work-for-free-while-the-owners-get-rich model works any more.
I appreciate the feedback and the connection! Clearly, the content creation models are quite different; our focus is pretty exclusively on hard news (for now); we summarize (and can rearrange) individual facts from many sources instead of making one narrative per story written by one person; and you don't rank the news in any meaningful way, as far as I can tell (do correct me if I'm wrong, but I don't see how).
We've got quite a few very motivated writers who find writing bitts to be fun, and they know they're working on a site that will be open content (complete with API) and thus a potentially forkable public resource. We're considering doing a profit-sharing system, but I'm worried about the effect that will have on the community. We haven't even hard-launched (the content is still behind a login), we have had minimal publicity, and yet we've got enough regular participants to have a decent evening edition. Relying exclusively on paid writers will prove a drain on resources and might pose a significant roadblock to allowing the site to scale. It's a good thing we didn't do that with Wikipedia...
Thanks Larry. I think breaking down the info into bits is interesting, it's something we have been looking into for other parts of our site (not news). BTW, weren't you paid to create the first 1000 or so Wikipedia articles?
I was Wikipedia's co-founder and first and so far only community manager/chief instigator. I did contribute a bunch of articles, don't know how many, while in that position. I wasn't paid per article.
If you don't pay people, you have to compensate them in other ways--especially in terms of community recognition and other evidence of the value of their work to the community/readership as a whole.
Our writers have added almost 25,000 posts on thousands of topics (a sample here http://newslines.org). So far, at $1 a post, that has cost $25,000. Some of our writers have made thousands of dollars. We just closed more investment are are building a revenue-share model that will give our writers even more money. People who get in early, and do consistent work will be able to make a lot of money.
"Wikipedia for news" might be a brief statement of the purpose of Wikinews, but it isn't an attempt to solve the problem I'm concerned to solve. Wikipedia made encyclopedias better by making a giant encyclopedia. Wikinews didn't make news better...in any way. What we want to do is actually organize the news, including the "long tail" of citizen journalism, into "small pieces, loosely joined" in the form of one-sentence fact summaries. The result will enable a scalable community to co-author a truly useful, complete, and giant selection of the news, without the noise and confusion that besets the news today. So that's the difference.
This I could get behind. If you were to strip the adjectives out of the stories (both literally and figuratively, it would not only remove some of the inherent bias, but also a lot of the noise. For example: "A police officer shot a man during an altercation over Lucky Charms in Sometown, California. The shooting is currently under investigation.
And that's it. You report the facts of the investigation and then the public has an unbiased view of the facts, without the subtle nuances that introduce bias. Such as, "A black police officer shot a Vietnamese man. Until the facts of the case reveal that race was a motive, then those details presumptively introduce race as a motive. That's the kind of journalism that is successful financially because it generates strong emotions, and thus viewers however it's subtly dishonest not because those facts aren't true, but because of the motivation behind why those facts where included.
This is Larry Sanger, one of the original founders of Wikipedia, but he disagreed with it very early and left to start a competing project, “Citizendium”, which went slowly and stagnated a few years back.
I'd prefer to say that it started with a bang (it really did) but slowed way down due to the unsupportable weight of trying to compete with the 800-pound-gorilla, Wikipedia. :-)
Also, isn't it harmful to divert effort that could be going into WikiNews to a different project? WikiNews already struggles from a lack of editors. Is this substantially different in some way?
So, everyone interested in this idea should put their efforts into a failed project. No alternatives, because that would be harmful to a project people clearly aren't interested in developing.
If wikinews isn't able to get this idea working, why is everyone else required to keep supporting it. Maybe it's time for an alternative.
If it's not working, then it needs needs to figure out why people aren't engaged. Otherwise, the natural progression is for people to move on to an alternative.
Yes, it's very harmful to divert all the effort everyone's apparently (not) giving to WikiNews to a different project that actually might succeed in dealing with information overload, starting with front page news.
Why not? If nothing else, the new news site is, itself, news, which could serve as a PR boost to attract new contributors. Wikinews is old hat, nobody is going to start contributing there now.
A new site with a fresh start has a better chance of widening its contributorship.
I really like the idea of summarizing various source in infobitt. But I think it still suffers from some points outlined in "News is bad for you – and giving up reading it will make you happier" [0].
E.G. not living in the states, half of the "top news" section is pretty irrelevant to me and the relevancy of the others seems to be based on the personal interests of the tech sawy contributors. I'm sure this gets better once more people are using it. But i guess having some noise is unavoidable.
The nice thing about this crowd approach might or might not be, that only provable facts are quoted and fewer false accusations are made. Thats mostly the reason why I stopped following news after the Utøya massacre in 2011 and later quitted my job at a news site.
Is there already an API one can fiddle around with? I think there is a huge potential in being able to use that data and hopefully feed stuff back. In my opinion, the problem with journalism today is not journalism itself but the distribution of content and the lack of choice how to consume it.
The news on the front page looks a little weird because a bunch of new people from Hacker News just arrived and ranked the news very quickly and sloppily. So it's OK. In a few days when things have settled down (??), you'll see it's much more plausible. As for being U.S.-heavy, we will of course have a U.K. version, other national versions, other languages...but that'll be post-launch and post-second funding round.
In the long run our platform will make it possible for you to get the news ranked according to your preferences and interests, which will be made possible by the fact that we break the news into facts that people can rank differently. So your friends' (and compatriots') rankings can be up-weighted and what you're shown will be different.
I know this isn't what you see on the site yet...but in 2001 and 2002 you wouldn't see what you'd have liked on the front page of Wikipedia, either. :-)
Sorry, no API yet. We will have one; we're open content. We're not even hard-launched. We've built what you see on a shoestring...
Less and less do I trust the wisdom of the crowds, and Wikipedia. I want expertise, not the consensus of amateurs and the ignorant. Look how well the consensus of experts, or at least professionals, works at HN: useful, better than alternatives, but I wouldn't bank on it.
A story, possibly apocryphal, about Richard Feynman: He gave a strong negative review to a textbook that the other reviewers endorsed. When confronted he said he might not be the smartest person in the world, but was he more intelligent than the average of a hundred people? Certainly!
Or from a Car Talk brain teaser: Do two people who don't know what they're talking about know more or less than one persion who doesn't know what they're talking about?
What about 100 or 1,000 people? I'm pretty sure they know less, as they create in their echo chamber greater certainty and additional untruths.
(I am overstating the case for effect; there is value in the aggregate factual knowledge of crowds, and they are not always ignorant.)
As I have said several times before, Wikipedia doesn't follow Surowiecki's rules for "wisdom of crowds" effects.
The Infobitt model has an advantage over wikis that will enable a wisdom of crowds effect. People contribute and vote independently of each other, not by having to agree on a single version of the content. Yes, there can be comments, and that will have a biasing effect; but what people say in comments is obviously not as important as their votes, which they can exercise independently of each other and comments.
Also, since the pieces of content are short and fungible, they are therefore capable of being subject to contests. We can submit competing versions of facts, and the best can rise to the top.
Finally, every fact in Infobitt has to be sourced, and the fact is a summary of the source; and (soon) we'll be able to compete to write the best summary. So, to be sure, we'll have bad summaries occasionally, but generally I think we'll have good ones, once we've built the community up some more.
One idea is to develop credibility and reputation over time by contributing useful facts. It's not raw crowd numbers but rather contributors with reputation which will yield results. Think of StackOverflow but for news instead of computer Q&A.
Wikipedia contributions are not averages of average people; they are written by those who choose to participate, not by a random sample of the earth's population. In some fields, Wikipedia is held in high esteem by many professionals.
In the wonderful event that Infobitt becomes a superior product and accordingly gains market share to the detriment of news sources, then we will of course know where our bread is buttered. Infobitt will suck without good sources; so probably we'll end up paying our sources ourselves.
A much higher proportion of news content than Wikipedia content is politicized, and Wikipedia functions worst, IMHO, for politicized issues.
How will you prevent users from politicizing your content? For example, will you exclude more politicized sources such as Fox and the Huffington Post? Separate editorial sources from straight journalism? What about government-controlled media such as Russia Today (RT)?
It seems like the content of politicized stories could be mostly politicized 'bitts' (where often both sides are deceptive and none of the information is valid), and it could merely represent the beliefs of whichever side has more dedicated contributors.
>Wikipedia functions worst, IMHO, for politicized issues.
It still functions better than most major news outlets do. Usually the only time I see an article that is very heavily slanted is when it the pet issue of some particular group or cause (e.g. https://en.wikipedia.org/wiki/Antisemitism_in_the_United_Sta... )
Its rules on sources aren't bad at all, it's just implementation which is a bit dodgy sometimes.
With the voting on GN being time and search traffic? I am not sure how the determination on GN ordering works, but the idea of human curated news seems an even more sensation driven principle than the current network news industry.
Even HN falls victim to sensationalism on a regular basis, and I presume most of HN's users are somewhat invested in the quality.
I am not meaning to detract from the effort, but the outcome for such a scenario seems pretty repeatable. Especially with the history of non-niche vote driven news outlets/sites, and their predilection to not be viewed as quality, definitive, or substantive.
Ever since I heard of this site, I've thought that (annotating the web, or parts of the web) is a great idea for a website, and they're executing it pretty well. It's not what we're doing, but it's still cool and my guess is they'll keep growing.
No, it's not. I use Google News when trying to figure out how to organize and summarize the news. Google News isn't very good at ranking news, and it doesn't even try to summarize the facts in the news.
Is this something Infobitt evaluated or do you mean they don't rank it well for your personal needs? Also, they do provide 1-2 sentence summaries, but nothing as detailed as Infobitt.
I happen to think Google ranks news surprisingly well, though not well enough to be my primary source (I doubt any machine or group of people will beat my personally selected and organized RSS feeds), but that's only for my needs. FWIW I don't logon to Google and block most tracking so I'm not sure they customize it for me.
Google news is likely based on algorithms and big data based on user searches and the # of times articles with keywords are getting selected. Nothing wrong with that, but far short of what an informed and active community can do. The two are complementary but quite different.
I have been contributing to infobitt for awhile now. The potential is for it to be a place where you get a rounded view of a story by multiple editors providing different facts. One option I hope will emerge is filters so I could see for example how the world news appears from an Australian perspective and then instantly flip and see it from a is perspective. That would be something that nobody has done as yet.
I've been contributing to infobitt now for about a year. I think the idea of breaking a story down into individual facts linked to sources is interesting. The game changer will come when we start adding value beyond that. There are some ideas percolating on that front. So I would encourage anyone interested to get on board and see where this takes us.
Folks already contributing hundreds of hours for free for some time.
I can't imagine someone doing free work and writing, ranking restaurant reviews, editing, citing, writing pieces of encyclopedia entrees or electing Presidents like in 2008.
Getting free contributions works for Wikipedia because it is not a business, has no ads and it's supported by donations. And because people get ego value from getting their contributions on a large, popular site. However, if the intention is to use free content to make the owners (and VCs) rich, then I am not sure that is a sustainable business model in the long term.
Yes, it's possible. I haven't seen a crowd-sourced news website with a focus on the community so much as Infobitt. The site's style is really good: the news is given to me in small pieces with easily digestible key facts. I can see a lot of people going to this site every day to know what's going on around the world.
Sounds like a nice idea but isn't Google News already doing a decent job on this? It's going to be hard to beat the speed of Twitter or Google News in my opinion but I'll be curious to see where this goes.
Not really. Google doesn't summarize facts. It aggregates articles into clusters, which is useful, but it doesn't save us much time. Google also does a rather poor job of ranking the importance of stories, and often has old news for a long time and breaking news relatively late (or later than a motivated community, like Twitter, can have it).
Google News' strength is ranking, not summarizing facts from different sources so you can avoid reading all the top 5 or 20 articles about a story.
And, philosophically, do you HN folks really want to rely on the MSM to tell you what's important? And likes, shares and troll-ridden comments on news stories?
Or do you want a democratized way to decide, collectively, what news stories and facts are important.
Don't forget the NYT, etc. screwed up on no WMD's in Iraq.
I was at a small group dinner with Judith Miller, ex-NYT, who's nice but she messed up.
I'm a happy user of InfoBitt. I went from skimming different news sources, such as Yahoo headlines, to strictly looking at InfoBitt for the top news of the day.
I'm pretty sure it isn't. People who claim the opposite seem to miss something important. Well, of course it depends on how to define "wikipedia for news", but there are several reasons why it is empty talking.
First off, wikipedia is all about data. It's really cool that it provides easy to use service as well, and that's the reason why it is somewhat more successful than OSM, but nevertheless, Wikipedia is the data. Newspapers, TV, now all these news portals are services. There is important difference between data and service.
Data is gathered and shared amongst us as people working for some great good, which is useful for us personally as well. I might event not like you, disagree with all your opinions, but as long as you can provide to that great work of ours some knowledge that I cannot provide (even if I'm not particularly interested in it) I welcome you. All that matters to me is that you are not lying here. And, as you can see, even in such (presumably) politically-neutral environment there is much disagreement and silly behavior, people tend to get personal, there're edit wars, forked projects like encyclopedia dramatica, because there obviously appears to be some content which isn't interesting for one community, but interesting for another. I don't know much about content of sites like knowyourmeme and such, but russian clone of lurkmore is actually a funny example, as many of articles there are about some real, important topic, about which article on Wikipedia also exists, but are composed in a much more harsh manner, without worrying about political neutrality, and often delivering some curious facts, so if you are interested in the subject you would probably read article both there and on wikipedia.
Service is something to be delivered. It must be on time, as "cold news" aren't even news anymore. It's about you providing me information I'm interested in even before I know I'm interested in it, so you should guess it (no matter if it's having good intuition or using machine learning). It's about it being provided in right amount, so I wouldn't stop reading before I get to the most interesting part (and never buy a newspaper from you anymore). It should be reasonably entertaining, so I would want to come back for more. That being said, service is kinda hard. And sadly I assume you don't want to work your sweat off just to please me, for free it is. So our little community-driven platform should be as useful for me, as it is for you. There are several easily deductible reasons why it is a problem, so I'll skip discussing them and get to the first conclusion: something that is about opinions and is equally useful and interesting for all participants isn't news service, it is social network. So if you think you are building news service I guess you don't understand what you are building, because actually you are trying to build one more social network. Lack of understanding what you are making is a problem by itself.
Second is empirical confirmation of the first, and is pointed out in other comments: we have plenty of services like that and services which are social networks in the first place (reddit, HN, even twitter for that matter) are more successful news platforms that specialized news platforms. And I don't even see any claims of how different form them it would be.
Third problem is as much as I don't like journalists, there are reasons for them to exist. They go to dangerous places and make photos, they use all kinds of shady tricks to find ugly and quite interesting story under plain-looking surface, they know who to ask, they know how to ask. They know what to tell to their consumers, they know how to tell. If you are building your own virtual newsroom without journalists you either need to use resources provided by real ones working for other agencies, which makes your own platform some aggregator like facebook or google adds, or, yeah, reddit, HN, Twitter, everything else. Or you just won't have anything (interesting) to tell, really.
Ah, my bad. I didn't sign up because I never sign up to i-don't-know-what and I didn't expect that I will find some "about page" after signing up, which seems to me like very natural thing to [do not] expect. Unfortunately I can't follow your link without signing up either, and service that behaves like that actually makes me inclined to go away and never turn back not matter what. So, no, I didn't read it, and I admit it very much might be the reason I don't understand what it is about.
But, still, what I said in the previous post are pretty general statements which I believe must be true to some degree for every single form of "wikipedia for news".
I suspect that after the site undergoes some further improvements, its content will be made visible to those without accounts (so that you would need an account to contribute but not to read).
But why you need people to log in to see about page? Just 1 static html page so I (not really me, because I spent a lot of time on that conversation anyway, but somebody, whatever) could decide if it's worth my time to sign up using real email account. It's, well, the point of about pages, to explain people what is that stuff they are looking at, and if they really want to go further. Scalability issues? That 1 static html page could be hosted anywhere, and, besides, if your servers aren't dying to host login page it wouldn't make very much difference anyway.
It's certainly something we can consider now that we've soft-launched. We've thought about it in the past, and now I guess it is a good idea to consider it again!
The site has not hard launched yet so you have to create an account to see the content. But I encourage you to do so. You will see how much different it is than Reddit.
