Wikipedia:Wikipedia Signpost/2023-12-04/Essay

Essay

I am going to die

Memento mori: Remember that we will die.

Here is an uncomfortable fact: I am going to die. I hope it won't happen any time soon, but it will happen, and it will happen to you, too.

On that day, will there still be editors here?

Replacing me

I firmly believe that I am not irreplaceable. However, I also don't think that I'll be easy to replace. Looking at the numbers, here's what we need to replace me:

100,000 people to create an account.
30,000 of them will make a first edit. (The other 70,000 will never find the edit buttons, or they'll open a page, see complicated code, and give up.)
20,000 of them will make a second edit. (This usually happens within minutes of the first edit.)
10,000 of them will make it to their fifth edit.
5,000 of them will make it to their tenth edit.
1,000 of them will make it to their 100th edit.
100 of them will make it to their 1,000th edit.
25 of them will make 10,000 edits.
5 of them will make 50,000 edits.
One of those original 100,000 accounts will eventually match my volume.

Trends over time

When we're gone, will there be anyone to take our place?

The number of active editors has been pretty flat for about a decade. It might occasionally feel like everyone's quitting, particularly if a wiki-friend disappears, but we're holding on. However, anything that affects retention – whether that's a change we control, like how we treat each other, or a change outside of our control, such as an economic recession – could destabilize our balance.

We get about four million new accounts each year. That means that, in a given year, we can replace about 40 of the editors listed in Wikipedia:List of Wikipedians by number of edits#1–1000.

Everyone on that list is going to die some day. There are 1,000 editors on that list, and we are getting enough new accounts to replace about 40 per year. That's only 4% per year. We can only maintain our current situation if the average editor on that list stays active for 25 years. If we lose promising new editors, we may not have any editors left.

Replacing you

How many newbies do we need to replace you? Check your edit count at Special:CentralAuth, and find yourself in this list:

If you are newly autoconfirmed (10 edits): 20 new accounts, 7 first edits
If you have made 100 edits: 100 new accounts, 30 first edits
If you are newly extended-confirmed (500 edits): 500 new accounts, 150 first edits
If you have made 1,000 edits: 1,000 new accounts, 300 first edits
If you have made 10,000 edits: 4,000 new accounts, 1,300 first edits
If you have made 25,000 edits: 10,000 new accounts, 3,000 first edits
If you have made 50,000 edits: 20,000 new accounts, 6,000 first edits
If you have made 100,000 edits: 100,000 new accounts, 30,000 first edits
If you have made 500,000 edits: 1,000,000 new accounts, 300,000 first edits

What to do about this

We need to see clearly into the not-so-distant future.

The first is: Please do not bite the newcomers. Remember not just that you will die someday, but also that you were once new, inexperienced, clumsy, prone to breaking things, unaware of any of the rules, and generally making mistakes and screwing up articles. Go look at that list again: That's how many clueless newbies the previous generation of editors had to tolerate to get an editor like you. If we want editors like us to be here when we die, we need to extend the same grace to the current newbies. We won't get another generation of editors by making them feel unwelcome, especially since our environment is more complex than it used to be, and the alternative outlets (e.g., social media and video sites) are much more available and attractive than they used to be. We have to compete for potential editors' time and interest.

The second is: Remember that when you are "defending the wiki" in the short term, you could be killing it in the long term. We need new editors more than we need the endorphin rush of insta-reverting an uncited but probably accurate contribution. We need new editors more than we need to make things convenient for us. We need new editors to feel like they can be successful more than we need to get rid of subjects with borderline notability. Yes, you are permitted to blank uncited content, and sometimes articles really do need to be deleted. We do get a lot of attempted spam and self-promotion. But if we don't want Wikipedia to die off when we die, we are sometimes going to have to WP:FIXTHEPROBLEM ourselves, instead of pointing fingers at the less-experienced folks who didn't get it right the first time. We are often going to have to encourage and support the newcomers. We need to thank every good contribution from a new editor. The options shouldn't be reversion or radio silence; the options should include frequent thanks and praise and encouragement and enthusiastically building on their work.

The third is: Create actual content. One of the problems with evaluating replacement rates according to edit count is that it's possible to rack up a lot of edits by doing nothing of real significance. A well-written article is a solid contribution to the world; changing the order that refs appear in a list is not so important. If you dislike unsourced articles, go source them. If you dislike outdated articles, go update them. But the main thing is: figure out which articles are missing or in significant need of expansion, and go create some actual content. Good content will outlive you.

← Previous "Essay"

Next "Essay" →

In this issue

4 December 2023 (all comments)

News and notes

In the media

Disinformation report

In focus

Comix

Essay

Discuss this story

These comments are automatically transcluded from this article's talk page. To follow comments, add the page to your watchlist. If your comment has not appeared here, you can try purging the cache.

Very well stated. — Maile (talk) 13:58, 4 December 2023 (UTC)[reply]
Agreed, this is an effective essay. Turning readers into editors is not easy, but we need the pipeline operational! —Ganesha811 (talk) 14:04, 4 December 2023 (UTC)[reply]

Actually, I'm one of a few editors who's made an unusual amount of edits without, some would argue, creating "actual content." In fact, there's been times I've stubbified articles full of peacocks, weasels, and other forms of puffery. Even though I'm easily replaced by someone who WikiGnomes and also content edits, I'd say that my edits are still valuable, since it frees up the time of content editors to actually add content. (I'd trust the author realizes this, having had conversed with her in the past, but I fear others may protest to the contrary.) --I dream of horses ^{(Contribs) (Talk)} 15:42, 4 December 2023 (UTC)[reply]

When the current generation of active editors passes on, their replacements will probably be a mix of:

The WMF eventually opened its purse and hired a real editorial staff and topic lead editors
AI tools matured and allowed these editors to generate sensible copy for new articles
The obsolete wiki markup is replaced with something more easily machine-editable
Anonymous editing goes by the wayside
Truculent editors are shown the door
New editors are recruited from academia and news sites
The Five Pillars are made into a nice needlepoint and displayed at WMF HQ

Or Wikipedia dies. →StaniStani 21:51, 4 December 2023 (UTC)[reply]

At that point, it would already be dead. What you're describing is not Wikipedia. Citizendium already tried something much like that, and well—it's been dead for a very long time. Seraphimblade ^{Talk to me} 06:11, 5 December 2023 (UTC)[reply]

(Ironically, even with all our values thrown away, the name would likely still be "Wikipedia", and this may become an issue one day. Forking is hard if the brand is stronger than the content and the community. "OpenOffice" to "LibreOffice", "MySQL" to "MariaDB", ... "Wikipedia" to anything else would be far more than just a technical challenge.) ~ ToBeFree (talk) 18:58, 5 December 2023 (UTC)[reply]

Predictions of the end of Wikipedia -- ssr (talk) 15:20, 10 December 2023 (UTC)[reply]

When AI tools mature enough to generate sensible copy for new WP articles, they won't have to. They'll just do that, custom made for the knowledge and tastes of the user, when desired. Meanwhile, at editathons I sometimes say, "No, I no longer think Wikipedia is too small, and seldom make Wikipedia bigger by adding good things anymore. Mostly I make Wikipedia smaller by subtracting bad things." It is perhaps an exaggeration, but I think it makes the point. No use expanding if we don't have enough maintenance workers checking for new bloat, spam, redundancy, lies, and other bad content from the many ignorant and the few malicious. Jim.henderson (talk) 05:17, 5 December 2023 (UTC)[reply]
The last paragraph sums up perfectly what I've learned about Wikipedia and how you should behave on it in my (almost) two first years of activity around here. Thank you! Oltrepier (talk) 09:28, 5 December 2023 (UTC)[reply]
Good article, but three things I'd add. Firstly the effect of SUL, Wikimedians on other projects especially other languages, will come visit here while logged in. They may even copy an image, but if they don't make an edit here they just count as a new account created here but not editing on this Wiki. My own account has gone live on 174 projects, about a sixth of the total, but on most of those I have zero edits and only on four do I have over a thousand, I expect many of you will be similar. Secondly, and probably much bigger, just over a decade ago we replaced a lot of vandal fighting with edit filters. So an unknown proportion of accounts with zero edits are vandals whose "contribution" was rejected as vandalism. OK not all of those filter rejections will be accurate, but those filters were thoroughly tested when they were first deployed. Maybe we need to review them now as the culture has changed and there could be a word which twelve years ago was almost always used in vandalism, but now it is the name of a successful song or Ukrainian village. Thirdly, and probably most importantly, the last decade or so has been the era of the smartphone. I have been told that despite my own personal failure it is technically possible to edit with a smartphone. But very few of our community are from the large proportion of internet users who only access the internet by smartphone, and in my opinion this is the biggest driver of the lack of growth of the community at least during the current era of the post 2014 rally, and probably for some years before during the latter part of the 2008-14 "death spiral" era of declining editing rates. "active editors" shows quite neatly the three thirds of Wikipedia's history thus far, the exponential growth of 2001-7, the increasingly misnamed death spiral era and our own normality, the post 2014 rally of 2015 to date. I no longer fret about this as much as I used to. I hope that one day an AI typo fixer will be trained on my contributions and take over such easily confused words as calvary/cavalry, pubic/public, preform/perform, manger/manager, minster/minister and causality/casualty, and maybe there are other tasks where AI can replace editors. But to really make a difference to editor recruitment we need to solve our smartphone problem. Maybe in the next few decades smartphones will go the way of punch cards and wind up car windows, and perhaps their successor(s) will be better editing devices. Maybe nothing much will change technologically in the next forty years and the community will then be rejuvenated by a generation of retirees who grew up with smartphones and don't see how they were ever a problem. I used to naïvely hope that the WMF would see the problem and get developers to fix it. Ϣere SpielChequers 11:12, 5 December 2023 (UTC)[reply]
The essay resonates with me, but the comments also make good points. It is a complicated matter. Add to the list of problems that much of the low-hanging stuff has been done, sources are often hard to find, and even when found, are often inaccessible, and a large proportion of knowledge is ineligible because notability cannot be shown. · · · Peter Southwood ^(talk): 05:52, 7 December 2023 (UTC)[reply]
There will be editors, maybe even active ones, just that if the average editor doesn't last here for 25 years, there can be times when there will not be active editors in the top 1,000 (also, there may be times when there will be no active editors in the top 1,000 if the average editor does last here for more than 25 years, however, the longer they last, the shorter that time becomes). Also, there are many unregistered editors as well, editing approximately every second. However, if there are not enough active and well-behaving registered ones, missing articles will be long missing. Alfa-ketosav (talk) 15:23, 8 December 2023 (UTC)[reply]
Well, I for one am trying to fill the gap by writing an article on a subject that I'm honestly surprised we don't currently have an article about. (It's the kind of subject that you'd expect a white, suburban, educated, middle-class male from the US would have already written.) I haven't written much of this article because I needed to do some serious research first (e.g., it's context in a larger subject) which required collecting reliable sources, & now I'm trying to figure out how to present the material. But given enough free time, I will get it written, for non-Wikipedia reasons. -- llywrch (talk) 00:41, 9 December 2023 (UTC)[reply]
There was a time when scholarly papers and sometimes books were typed on word processor screens with 20 lines of 80 characters, and apparently there are now people who feel comfortable using a palm-sized screen to write a moderate sized Wikipedia article. My guess is no, content creators in this field and others will for a long time prefer a bigger canvas than is used in consuming the product. Certainly the software could be better; it's easier to write long Quora posts on a little touch screen than short WP articles and goodness, why can't Wikidata statements and Commons categories be easily entered in the field? Anyway I've been teaching newbies this year at a slower rate than before the plague, at fewer editathons. On average they're at least as bright as the ones before, but yes, their lesser numbers support the concern that as we old-time copyeditors slip into our commas, we won't be replaced. The Admin corps is already in such a situation. Jim.henderson (talk) 14:25, 9 December 2023 (UTC)[reply]
I would add to the second point of "What to do about this" to note that it's very discouraging to add good faith content and then find it's reverted by a Page Guardian When that happens I find myself wondering if I want to waste my time arguing with a trigger happy Page Guardian and eventually start looking for useful things to do elsewhere.

Doing something to discourage the behavior of Page Guardians might help to keep more of the productive editors that we do have. Perhaps there could be a reversion patrol to jump in with an independent view on any reversions which are other than reversion of obvious vandalism or the other instant reversions for policy reasons (libel, copyvio, etc.). 2A01:CB19:599:F00:212:4B42:693:9588 (talk) 20:33, 9 December 2023 (UTC)[reply]

I like your idea of reversion patrol. I have been thinking about that for a long time. We now have rollbacker, patroller, etc. But nothing like reversion patrol. Power is not balanced. And the reward system is skewed. This needs to be fixed. --Dustfreeworld (talk) 18:03, 14 December 2023 (UTC)[reply]

Except not really, as many IP editors do edit. I've made four edits in mainspace, more than 80 percent of registered editors. So it wouldn't need quite so many as it sounds. IP editors are rarely super prolific, but still, that's useful. 71.112.180.130 (talk) 15:28, 10 December 2023 (UTC)[reply]
Sheesh, if the numbers for editors are this bad, I shudder to think about the numbers for admins. We're down to how many monthly-active admins these days? How many successful RfAs per week? Somebody should do the math and figure out exactly how many years we have before we'll have to start posting to meta:Steward requests/Miscellaneous. --N Y Kevin 23:53, 12 December 2023 (UTC)[reply]
A quick little regression on the number of active admins reported at Wikipedia:List of administrators/Active on today's day for the last 10 years suggests it'll be about half a century until we reach zero. WhatamIdoing (talk) 00:32, 13 December 2023 (UTC)[reply]
Some of our youngest current admins might still be around in half a century. But we go beyond any sort of reasonable crisis point long before we just have 1 admin. Aside from the number of admins needed to keep the deletion process running, we always need at least one admin at AIV. With fulltime staff it takes five people to have one person on duty 24/7 - 6 if you allow for holidays and sick leave. Doing that with volunteers needs a lot more, especially if you aren't running some sort of rota but just relying on having so many volunteers that there is always one around. That isn't an argument for a rota, if we ever get to the point where we need one, we immediately turn adminship into more of a chore that is harder to recruit for or retain volunteers for. And given the size of EN Wikipedia, I don't see we can do this with stewards from Meta.

But to answer Kevin's point, the number of active editors has been broadly stable for a decade or more, our problem is that fewer and fewer of them are volunteering to be admins. Ϣere SpielChequers 09:08, 13 December 2023 (UTC)[reply]

Explore Wikipedia history by browsing The Signpost archives.

Home

About