User:Kate/WhyWikiPediaWorksNot

~~Or: “Why Wikipedia is doomed to failure”.~~

Please note: if, after you read this, your reaction is “well, if you don't like it, go away“, please understand that I am not not saying people should not contribute to Wikipedia and I am not saying Wikipedia has done/will do nothing useful. Wikipedia is useful, I do contribute to it and I do not plan to stop either reading or contributing. I merely believe that, in the future, Wikipedia may not be quite so useful.

Factual errors

Wikipedia has factual errors—I don't think anyone can deny that. Any written work of this size does. The question is, “how many?”; and the answer: no–one knows. Even if every article is created in good faith and entirely accurate—and that alone is dubious—how many errors have been introduced since then, whether by anonymous vandals, or POV–pushers, or simply well–intentioned but misinformed normal users?

At first, this wasn't a problem. Vandalism was quickly noticed and corrected, factual errors or omissions were researched and the articles corrected. But now, after only 3 years, the sheer number of edits in recent changes, the number of articles on an awesome variety of topics, is just overwhelming. No–one can possibly verify every single edit; no–one has the knowledge needed to spot—and correct—the factual errors that are introduced every day. Wikipedia is being crushed under its own weight. Its openness, the very feature that makes it so successful, is fast becoming its own downfall. Already we see vandals who deliberately introduce factual errors into articles; changing numbers, dates… most of them, perhaps, are caught, but how many aren't? Again—no–one knows.

In time—one year, two years, four years—Wikipedia will be worthless for more than anything other than a basic overview of a topic. Citing it as a reference source for facts will be laughable. As an encyclopedia, Wikipedia will be a failure.

Internal politics

Wikipedia is drowning in politics. Requests for arbitration, requests for comments, what sysops are, or aren't, or should do, or shouldn't do; what a vandal is, or what a troll is; whether such–and–such a user can be banned or not, whether this or that article can be deleted…

All this contributes nothing to Wikipedia. It distracts editors from real work—writing articles—and drives off outsiders and new users, not to mention existing editors. And this is increasing, not going away. Every day there are more arguments; more vandals, more trolls. This is arguably something that could be resolved by change; a change in policy, a change in process… but where is that change? There are no wonderful new ideas on how Wikipedia should be run. Any unresolved dispute is solved by “ask Jimbo”. This is not scalable. If Wikipedia doesn't take a long look at solving these problems, they may well end up contributing more to Wikipedia's death than even gross factual errors.

Inherent weakness of Wiki

This is one of the most obvious, and yet—so far—one of the least troublesome problems with Wikipedia. But although the kind of person who would deliberately vandalise Wikipedia in a major way might be rare, they do exist, and it's only a matter of growth before they begin to be a problem. What happens when DDoS networks are created to perform vandalism? Restrictions on new users, blocking of users or IPs, or entire networks, and manual reversion can only go so far. The only argument against this appears to be “well, it hasn't happened yet”.

In such massive attack, probably the most logical way is to revert everything to the nearest full backup and then implementing CAPTCHA for non-logged-in users. Also the number of edits per IP per time per degree-of-trust (perhaps measured by the number of their edits being labeled as vandalism (by different people, to contain personal vendettas)) of any given logged-in user can be throttled in some sane way. CAPTCHA eliminates using bots, the edit limits may throttle any individual malicious user beyond the level manageable manually. Could this work? --Shaddack 5 July 2005 00:59 (UTC)

This inherent weakness will end up imposing an effective cap on the growth of any Wiki—once it reaches a certain size, concerted vandalism will devastate it, and Wikipedia is no exception.

—Kate | Talk 03:51, 2004 Jul 5 (UTC)

Comments

Dysprosia on politics

I don't know about the factual accuracy comment, but I agree with you 100% on the "Internal politics" thing. Dysprosia 04:00, 5 Jul 2004 (UTC)

I was actually not too sure about politics. It's possible that at the worst, it'd just mean editors come, edit for a few weeks and leave. Inefficient, but not fatal. On the other hand, though, it's certainly bad for PR, and when Wikipedia relies on donations and grants for its existence… —Kate | Talk 04:13, 2004 Jul 5 (UTC)

Shibboleth on facts & community

You have identified two problems with wikipedia, but neither of them come close to implying total collapse of Wikipedia, IMHO.

Misinformed normal users do indeed inject a lot of factual inaccuracies into new articles. Indeed, new Wikipedia articles are often bad. However, the key point is that uninformed users are not going to change the facts that have been written by experts. Thus this group will not decrease the quality of Wikipedia with time.

POV warriors are generally extremely blatant in the changes they make. They will replace entire sections to reflect their extreme viewpoint. It is easy to detect and revert POV warrior edits.

Vandals. Now you have noticed that 95% of vandals do things like replace entire pages like "blah blah blah penis". These kids are no danger to Wikipedia. The real problem is users that make false factual changes. Now to do this, you have to be genuinely malicious, and explicitly desire the downfall of Wikipedia, not just act on a silly graffiti impulse. There are very few people this nasty. This is proven by the rarity of truly malicious computer viruses, despite how easy it is to make one. And if they do a lot of vandalism like this, one of their edits will be detected and then the rest of the edits from their username/IP address will be scrutinized. Finally, do you really think anyone would have the energy to insert subtle factual errors into thousands of articles? Who would do such a thing? This is very much a contained problem.

Finally, it is true that Wikipedia will never be 100% accurate. But what is? There are so many errors in mainstream news sources, let alone the rest of the web. Yet people still use regular web pages for information. Wikipedia will still be much better quality than most web pages even if there are some scattered factual errors. It may never reach the accuracy of the Encyclopedia Britannica, but this will be more than made up for by the *much* larger number of topics covered here. Perhaps it will not be referenced in scholarly articles, but it will be (and already is!) a very useful resource for general use. You realize that on hundreds or thousands of topics, Wikipedia is already the top Google hit?

Second: bureaucracy. It's natural that there will be more bureaucracy as the userbase gets larger. It takes a lot of time from the editors, yes. But a lack of people writing articles is hardly the problem. And it hardly affects new users: they will never notice unless visit the arbitration pages. No collapse here.

To me, the only real threat to Wikipedia is the danger of a Kuro5hin-style collapse caused by a change in the community. What happened to Kuro5hin is that an influx of bad new users came, overwhelming the old community and drastically lowering the quality of the discussion. Wikipedia needs to take care to properly assimilate all new contributors into its standards of quality. But fortunately, it seems to have been doing this very well so far. I am also worried about running out of money (the servers are already too slow), but hopefully as Wikipedia gets a higher profile the donations will increase until it's no longer a problem. It only takes one billionaire donator, after all.

Considering the amazing success of Wikipedia over the first three years of its existence, I don't see why you're so pessimistic. --Shibboleth 04:28, 5 Jul 2004 (UTC)

There are so many errors in mainstream news sources, let alone the rest of the web. Yet people still use regular web pages for information.

Yes—but isn't Wikipedia more than just a collection of web pages? I'd like to think Wikipedia could be an encyclopedia, an authoritative reference on almost any topic. Perhaps this is simply not possible for a Wiki.

[If] they do a lot of vandalism like this, one of their edits will be detected and then the rest of the edits from their username/IP address will be scrutinized.

In some cases, yes. But how many? If someone finds one error, corrects it, and goes off to do something else, it's likely enough to be lost in the history and never noticed by anyone else. Or a piece of POV or error may be rewritten, rephrased, copyedited, to the point where it's impossible to find where the original error was introduced.

There is also a delay between the error or vandalism being introduced and it being found and corrected. In most cases, one would hope, a small delay; but what about this vandalism, left uncorrected for a month, and found by a journalist? (From WP:VP). How many errors of this kind are there at any one time? Let's call these uncorrected errors the error corpus, since fancy terminology is so fashionable. As the size of tthe error corpus increases, Wikipedia's usefulness decreases. How long can the size of it be kept in check? —or even measured?

You realize that on hundreds or thousands of topics, Wikipedia is already the top Google hit?

Yes. That's how I found Wikipedia, actually—via pulse-code modulation, IIRC. But that isn't necessarily a good measure of quality, only popularity. And I hardly need to say that popularity is not proportional to quality in this day and age…

And finally, politics might not have a direct, measurable effect on articles, but it does cause editors to leave, and it does make Wikipedia look bad. And most importantly, it fractures the community, the very thing that Wikipedia relies on. Just look at Wik for an example. —Kate | Talk 05:08, 2004 Jul 5 (UTC)

I agree that the delays are an issue, indeed probably the main issue as far as errors goes. But your example of "Chesapeake, Virginia" is somewhat unfair, as this is one of the autogenerated pages that's almost never visited by anybody. What really matters is not the absolute size of the "error corpus", but which percentage of the pages on Wikipedia are errorfied. On this measure "Chesapeake, Virginia" barely matters, since it would be visited by one person per year at most. You'll notice that the most popular pages, such as World War II, are kept of very high quality. This is what really matters. I agree that there will be errors, and many of them, but it won't be to a disastrous level unless the quality of our main contributors (which are currently for the most part highly serious and educated) collapses.

I don't know, someone like Wik is bound to cause problems no matter what reforms we make to the system. I'm a new user and wasn't around back when there was the Wik battles, but I understand he vandalized pages when he couldn't get what he wanted.

Anyway, I agree that Wikipedia has its problems, but I mostly object to your claim that it's "doomed to failure". The evidence doesn't indicate that Wikipedia is heading towards doom at present. It's possible that it may collapse in the future, after all this is all a grand experiment and it's too early to tell --- but the indications are currently that Wikipedia is heading for success, and indeed is already quite successful.

I don't know, personally I don't care if Wikipedia is ever considered a "true" encyclopedia. As long at it becomes widely popular and useful to lots of people, I'll be happy. It's already more useful than Britannica, not because it's better but because Britannica is non-free and there's no way to read it unless you pay for it, which most people won't do. I've read a thousand times more encyclopedia articles since I started visiting Wikipedia than I ever did previously. --Shibboleth 05:36, 5 Jul 2004 (UTC)

I agree strongly with that point, of course we will have inaccuracies in less known articles, however articles which are being constantly watched by a group of people who are very well known to the subject will remain top-quality. --Ævar Arnfjörð Bjarmason 20:15, 2004 Jul 6 (UTC)

David Gerard on disruptive technologies

I propose a new slogan: Wikipedia: Even More Doomed Than Apple.

If you get three people in a room you have politics.

Look at it as a disruptive technology. These start with being better at some niche then expanding from there - David Gerard 12:28, 5 Jul 2004 (UTC)

Yes, but has Netcraft confirmed it?

I agree with Shibboleth, I think only about 10,000 entries accounted for most of the traffic. When we reach 10,000 active users, you'll only need each user to watch one of those articles (distributed watchlists) to combat perceptions. I'd like to see a person go unnoticed and mistakes go uncorrected at any large scale (i.e. more than a 1000 articles in 300,000+, not part of those top 10,000)

On a side note, Wik never vandailised in the main namespace (i.e. articles) as far as I know. Dori | Talk 19:21, Jul 6, 2004 (UTC)

The word you want is "beleaguered". -- Cyrius|✎ 19:34, 6 Jul 2004 (UTC)

Charles Matthews on other problems

None of those three would keep me awake at night. Certainly compared with (i) servers (ii) determined litigation (iii) the good editors getting fed up of the 'free riders'. Charles Matthews 20:27, 6 Jul 2004 (UTC)

Itai on politics

For my part, I'll take vandals over politicians any day. Politics - be it the forming of cliques or the piling of regulations - is a problem. The former is less a problem, as it is assumed would-be politicians would burn themselves out. The piling of regulations, which was not pointed out above, is, especially as there is currently no mechanism for "forgetting" regulation. I suppose one will arise eventually. Anyway, doomed is by no means the correct term, and even if it were, let it be. At the moment, I find myself increasingly unable to browse the web without popping into Wikipedia to clarify something, which basically means that Wikipedia 2004 is already quite good. At the very least - if Wikipedia disappears tomorrow - we've released quite a lot of quality material into the public domain, where it may live forever. GFDL! -- Itai 21:06, 6 Jul 2004 (UTC)

Wapcaplet on replies to common objections

Lady L.I., have you read Wikipedia:Replies to common objections? I'm not sure that hanging around Wikipedia for a mere month is enough time to accurately forecast its downfall. To date, Wikipedia has surpassed even our more optimistic expectations. As others have noted above, I don't think overall factual accuracy is in jeopardy; providing an anomalous single vandalism that wasn't caught for a month, out of an article body of over a quarter million, is hardly compelling evidence. How long would it take to have factual inaccuracies and omissions in Britannica corrected?

And what do you hope to accomplish by forecasting Wikipedia's doom? Since you seem to believe that it is dead already, why hang about and discuss it? If you are interested in helping us overcome these problems, it may be more productive to think of solutions, rather than simply thinking of the ways in which it may fail. -- Wapcaplet 23:28, 6 Jul 2004 (UTC)

Yes, I have read replies to objections. Think of this as something more specific—reasons I think Wikipedia might disappear entirely, rather than objects that might detract from or limit its success.

I don't think overall factual accuracy is in jeopardy; providing an anomalous single vandalism that wasn't caught for a month, out of an article body of over a quarter million, is hardly compelling evidence.

We'll see. That particular vandlism, obviously, (and as pointed out by someone else) would have been caught quickly if it weren't in an automated Rambot article. However, it was also a particularly obvious vandalism. What about more subtle things, from insertions of POV to changes to dates & statistics? Will those all be caught so quickly?

If you are interested in helping us overcome these problems, it may be more productive to think of solutions, rather than simply thinking of the ways in which it may fail.

Yes, thinking of solutions would certainly be nice. However, as you say, I've only been here a month (slightly less, I think), and I certainly don't feel I'm better able to suggest solutions than the many experienced people who are currently discussing, for example, modifications to Recent Changes.

Don't think that I spend all day thinking about how Wikipedia is doomed to failure :-). This was just a quick ponderance I wrote while bored one day—take it as you will. I certainly don't believe it's “dead already”, as evidenced by the fact that I'm still contributing edits.

—Kate | Talk 19:51, 2004 Jul 7 (UTC)

1) The ability to forecast wikipedia developement is not necessarily related to the time the forecaster is a wikipedia user. 2) Before thinking about solutions one probably sees problems. Writing them down increases other people awareness of the problems - and might let them think about solutions, multiplying the soluition search activity. Tobias Conradi (Talk) 21:01, 23 July 2005 (UTC)

CSTAR on dubious topics

Though I am concerned with the accuracy issue, I am most bothered by the enormous number of articles covering dubious material. Pokemon dolls? I just clicked random page and got room. Now room might be a useful entry, for example comparing rooms in Chinese culture with rooms in Spanish culture; but this can quickly spiral out of control. Another entry of dubious value world city; motherfucker seems to fall under that category, but in fact does have some interest. Similarly there is qubit and pure qubit state. Also the self-promotion pages are real annoyances as are the self-promoting references. (Also I have entered pages for living practicing academics and have regretted it, since they may be mistaken for self-promotion on their part) CSTAR 19:37, 7 Jul 2004 (UTC)

I don't actually see these as a problem. They don't detract from existing articles, and if there's something to say about them, why not say it? While they may be annoying, or take up space on the servers, I don't necessarily think they're contributing to Wikipedia's downfall. —Kate | Talk 19:41, 2004 Jul 7 (UTC)

If there are 3 million articles, quality control is much harder with the result that it is much more likely that a random query will end up with something worthless. Although this is not quite the same problem, I just noticed that in the article on Gaussian processes there is a link to prior (intended to mean prior distribution in Bayesian analysis) but in fact points to priory. CSTAR 20:14, 7 Jul 2004 (UTC)

World city looks encyclopedic to me (and note: I didn't write it). :) -- Stevietheman 13:04, 8 Jul 2004 (UTC)

Mackeriv on perfection

By reading this page, I can see you have a couple of arguments. Arguments that can be reduced to a main subject, the real fact responsible for your pessimistic opinion towards Wikipedia. In sum, you fear that this encyclopedia will evolve in a so intense way that, in the end, it will be no longer a reliable source of information, due to various reasons. You fear that vandalism and misconceptions will take over articles, which would eventually conspire for the end of Wikipedia. You fear that us, helping members of this project, will start to disagree with each other's views, something that will end up damaging this site's content. You've also been moving efforts into making every single article to follow a "pattern". You want them to look and feel the same. You want to reach perfection.

Of course, nothing is perfect. I don't expect or want Wikipedia to become 'perfect'. However, I would like to see it being 'acceptable'—and whether it can manage that is what I'm worried about.

Am I right in all of that? If that's what you really believe in, I am sorry to tell you, Wikipedia is already doomed. If you really think those issues will be capable of throwing this work into chaos, then, they really will. It's over. But there's something you haven't noticed. This is the Internet. You can't control the Internet. The cyberspace is a double-edged sword. While it is an infinite source of information about everything you can possibly think of, you can't completely rely on it. In that case, what really counts is the way this info gets to you. If you think of the Internet as a whole, you'll see that information is scattered everywhere. If you want to take advantage of the Net, you have to know where to find the info you want... and you also have to have the right skills in order to be able to select what is good for you and what isn't.

The way I see it, Wikipedia never was a 100% reliable source of information, and it will never be. No matter how many armies you recruit to control and standardize it, you'll never be able to make it completely trustworthy. It is impossible. However, a good thing about Wikipedia is that it's a place where people are able to gather a nice amount of information about zillions of different subjects. This is not supposed to be like a real encyclopedia, for a large number of reasons I shall not mention here. If it can't be considered like a regular encyclopedia, you can't expect it to behave like one either.

Think about this. If there are many people interested on turning Wikipedia into a pile of rubble, you can be sure there'll be an equally good number of other people interested on making it a nice encyclopedia.

Really? I'm not sure that's necessarily true. It remains to be seen whether the 'good' users can outnumber the 'bad'. —Kate | Talk 19:51, 2004 Aug 2 (UTC)

While not all of them can agree on everything, their mutual interests will be enough to keep things up. In my opinion, there's no reason to think that Wikipedia will be pretty much ruined in three or four years. I say it'll live. Dunno for how longer, and its form will probably be much different in the future, but I'm sure it'll still be there for quite a while. – Mackeriv 03:22, 9 Jul 2004 (UTC)

Isomorphic on accuracy

On the accuracy issue, I often repeat an interesting analysis I once heard from another Wikipedian (although I've unfortunately since forgotten who.) Basically, it was the idea that the importance of keeping a page accurate and high-quality is proportional to the number of people reading it. However, the number of people looking at the page is also the number of people who will potentially catch errors on the page.

So you end up with a situation where pages that people actually read are kept clean and relatively error-free. Error-prone or otherwise poor pages are the ones that no one is reading anyway. The percentage of readers who come to Wikipedia and get accurate information is much higher than the percentage of pages that are accurate.

To this, I will add a comment from my own experience using Wikipedia. In many cases, it's important just to know that two things are connected, or that if I'm interested in subject A, I should be also looking for information on subject B. This is especially useful if I'd never heard of subject B before. Here, accuracy isn't an issue at all, since "fact-checking" the existance of a connection between two things is usually trivial once the connection is suggested. Isomorphic 02:27, 31 Jul 2004 (UTC)

I think Wikipedia will be useful for a general overview for as long as it's around—but does that qualify it as an encyclopedia? Perhaps Wikipedia isn't destined to become an encyclopedia, but rather something new and different.

It seems reasonable that often-read pages would be more accurate, but it also seems a shame to lose one of the strength of wiki; that its contributor base is so large that over time it's likely to accumulate articles on many esoteric or niche topics, which fall into decay after their initial contributor(s) leave. Where is the line between 'accurate' and 'dubious' drawn? —Kate | Talk 02:46, 2004 Jul 31 (UTC)

As the number of readers increases, even pages on relatively obscure topics will be read fairly often by knowledgeable people. My personal hope is that some day Wikipedia will be such a universal reference that all knowledgeable specialists will feel obliged to make sure their field is covered accurately and completely. Isomorphic 05:04, 1 Aug 2004 (UTC)

That's an interesting point. Do you think Wikipedia will reach that point of growth before its hypothetical failure in that area from factual errors or other problems? —Kate | Talk 19:51, 2004 Aug 2 (UTC)

I just stumbled across this discussion, which strikes to the heart of what I've been finding as a potentially insrurmountable issue: accuracy. If Wikipedia is not to be accurate, then it's a large pile of random statements, some true, some false, some neutral, some so biased you can't tell what the NPOV would be. In which case, one has to ask "What's the point?"

In addition, though, there's the problem of inaccuracy in obscure branches of knowledge. It's contended above that the less frequently an article is viewed, the less important it's accuracy. I say that exactly the opposite is true. The more frequently an article is viewed, the more likely it is that some helpful editor will notice an inaccuracy and fix it. But if an article is about some field of study so obscure that it's viewed only once a year, and it's inaccurate, that one person in a year with enough knowledge to go looking for that article just found wrong information. If it's an expert in the field, he's gained a bad impression. If it's no, then it's even worse, because the reader has no way to tell that something he has no knowledge of is wrong.

Just my thoughts... Kenwarren 02:24, Aug 5, 2004 (UTC)

JamesMLane &c on factual errors

I agree with much of what you say. Uncorrected blatant vandalism is not, unfortunately, confined to automatically generated articles. An example that I found by chance: The article on the Scottsboro Boys referred to them as "the PePe Boys" from May 4 until July 17.

On the bright side, while you're correct that the number of problem users can be expected to rise, we can also expect an increase in the number of sincere contributors who are on hand to deal with the problems. For responding to vandalism, blatant POV, horribly mangled writing and extreme sloppiness, I suggest that the ratio between the two groups is what counts. I see no reason to think that the ratio will worsen. Perhaps we could improve it by being quicker on the trigger to block people.

It depends whether the ratio of good contributors rises proportionally to the number of vandals. I'm inclined to think that in the beginning, there are relatively few vandals and many good users, and over time the number of bad users rises. —Kate | Talk 19:51, 2004 Aug 2 (UTC)

The more serious problem is the factual error, whether introduced by subtle vandalism or by well-meaning mistake. To me, one big frustration is that people don't have to routinely justify their edits. Suppose I see that a particular date has been changed from 1939 to 1938. Unless I already happen to know the correct date, what am I supposed to do? Someone thought s/he knew it well enough to assert it was 1939. Someone else thought the same about changing it to 1938. Unless I'm willing to invest the time to research it independently, I'll probably just let it go. The natural result will be many instances in which Wikipedia makes an inaccurate assertion because someone plugged in something s/he thought was correct and didn't bother to check it. One solution, though it would impose a huge burden on participants, would be to require that all factual assertions be "footnoted". It could be an actual footnote where appropriate to the article, but much more common would be some sort of associated comment on the Talk page. Then there would at least be some prima facie indication that the edit wasn't pulled out of someone's garbled memory of a history class ten years earlier. The enforcement mechanism would be the community understanding that an unsupported edit can be summarily reverted. JamesMLane 10:45, 1 Aug 2004 (UTC)

Actually, the RC patrol exists to check edits like this. I'm not involved so I don't know much about it, but I know that if you see an edit you are suspicious of, you can let them know and they'll try to verify it. Isomorphic 00:10, 2 Aug 2004 (UTC)

I appreciate your letting me know about that feature. Nevertheless, it doesn't really address the concern I'm raising. I'm talking about the many, many edits that give me no reason to be suspicious -- except that they show that at least one Wikipedian was wrong. On occasion I may have some relevant knowledge myself. On other occasions, something may set off my BS detector and I'll pursue it further. But most of the time, necessarily, I let it go. JamesMLane 10:04, 2 Aug 2004 (UTC)

Mozzerati on technological escalation

Talking as a relatively new Wikipedia person, I think that some of this isn't such a problem since I come in and know, quite well, a specified area and set of pages surrounding it. Even very subtle wrong edits such as the recent ones reverted on GSM (one character change, between existing GSM versions, both of which have maximum power levels) are picked up easily by specialists in the area and reverted quickly (discussion about why the revert is correct may take longer). Generally, wrong edits just inspire some extra explanation. As long as my experience, as a new user, is true for the other new users, then the more editing there is the more correction is going to go on.

having said that, there's something missing technologically. For the pages I'm actively editing I would really like to easily see the sum total of changes since the last version I checked. Something equivalent to "what's new" for each page. I imagine then being able to click on an aknowledge button and next time I checked, I would see the changes relative to that newer version. Even with the easy comparison between distant versions, it's difficult to remember which was the last edit which was definitely okay. The database requirement for this would be a date stamp on each watchlist entry and programming would, I guess, need a couple of new PHP pages..

The important point here is that the task of checking new edits should scale with the number of edits, not with the number of existing pages (pages which haven't been edited shouldn't need examination)

other ideas for technological escalation:

measure views on diffs pages by (consenting?) logged in users; present a page with the least viewed diffs; ensure that most changes get checked by some people
logged in user is counted as "stable" after at least one week and at least 10 edits (minimises throw away accouts) edits by newer users are specifically tagged
two view wikki; public default is last edit by logged in user; logged in default is diff page for whole article with an edit was okay button
two view wikki applied to specific IP addresses (e.g. all AOL addresses cause two view pages); after editing, these users get a speficic explanation why they are in two view mode and an invitation to complain to their ISP and log in.
two view wikki applied to specific pages as a form of partial protection
provide a "dodgy content" button which any user can press to complain about content (without having to know much about wikipedia); check which pages come up most often; if being misused too much, provide an explanation page they have to press okay on (just because you don't like it doesn't mean there shouldn't be an article about it). Where appropriate suggest editing rather than complaining.
use public key crypto based cookies to create sub accounts of IP address accounts; warning visible during edit (and check box to ignore cookie); instructions given for clearning the cookie; many IP
introduce a proper system of Wikipedia:references so that we can also refer to paper texts. I suggest simply expanding the use of URLs to cover also paper texts through e.g. ISBN URIs etc.
introduce a system of hidden (by default) references so that there can be many many references in a text without detracting from its flow. Then encourage deletion of any material which someone questions, but which noone can provide a reference for. Similar process to votes for deletion allows warning of these planned deletions.

etc.

BTW. if any of these are already implented then you count that as a question about how to access these features :-)

some of these are definitely (slightly) harmful to openness, so use should be minimised, but the point is that right now wikipedia is quite okay, and in the case there start to be problems there's a long way that can be gone to make it more stable if the need arose. Most of these take some programming effort, so won't probably be done until there's a real need, but the fact that there are such options available can give hope for the future.

Mozzerati 09:01, 2004 Aug 5 (UTC)

Others

This looks very similar to claims that Capitalism is doomed to failure. Gosh, there's a lot of accurate information in Wikipedia? Who knew? Gosh, where the information is disputed the reader gets to know! Who knew? Gosh, its growing larger and more accurate all the time! Who knew? Guess there's something to the concept after all... - Tεx τ urε 21:14, 6 Jul 2004 (UTC)

*Rolls eyes* I hate pessimism much, much more than anything else. It must be all the Star Trek happy-endings. Rock on with that emo tune, Lady L.I. Good luck, and good day._{Peace Profound!} --Merovingian✍Talk 05:34, Jul 5, 2004 (UTC)

Well, it's not so much pessimism as speculation :-) The sub–title is intended to be slightly humourous. —Kate | Talk 06:40, 2004 Jul 5 (UTC)

Very interesting. I almost share your views. --Rrjanbiah 04:50, 8 Jul 2004 (UTC)

I don't think Cassandra was talking about things she found somewhat annoying, like potholes in the streets caused by the chariots. Where, by the way, is the evidence that things are getting worse? I sometimes find old articles that are in obscure places - they aren't as good as typical new postings from mid-2004. Charles Matthews 18:20, 8 Jul 2004 (UTC)

I agree with the problems that Lady Lysine Ikinsile has brought up. I think that Wikipedia has tended to vigoursly argue that it does not have problems whenever someone says that it does. Since Wikipedia in general works pretty well, there must not be too much of a problem. That doesn't mean that there are no problems in Wikipedia. I have started the page Wikipedia:Unsolved Problems as a place to discuss the problems that wikipedia still has. Jrincayc 19:30, 11 Jul 2004 (UTC)

I think that as wikipedia grows and becomes ever more popular it will bring it to the attention of people with more and more specific knowledge. They will tend to correct the errors, and wikipedia's accuracy will improve. I actually think that it may exceed the accuracy of printed books; including Encyclopedia Britanica; printed books are much harder to correct.

-Wolfkeeper

Factual errors

There is no evidence of this - in fact isn't your hypothesis unprovable?
The more users, the more vandals and the more anti-vandals. Sure, "No-one can possibly verify every single edit", but Everyone can.
There is no reason to think that popular or respected articles will decrease in quality, since many people are watching them.

Thus, Wikipedia is definately not going to suffer from the downfall you predict. You could say that average quality will decrease slightly as the number of articles balloons, but I don't think that's a problem. ··gracefool |☺ 03:15, 1 Sep 2004 (UTC)

Though you definately have a point with "Internal politics". Just look at the explosion of VfD... ··gracefool |☺ 05:10, 16 Jun 2005 (UTC)