April 3, 2024 • 37 mins
Wikipedia, a non-profit encyclopedia that anybody can edit, remains one of the few trustworthy and reliable websites left online. Ed Zitron is joined by critic, researcher and 18-year veteran of the Wikipedia editing community Molly White to discuss how Wikipedia actually works, and why you should care about its future.
See omnystudio.com/listener for privacy information.
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:02):
All Zone Media. Hello and welcome to Better Offline. I'm
your host ed zetron. Wikipedia is a free online encyclopedia
used by over a billion people a month, one that,
(00:24):
despite being maintained entirely by an army of free and
at times anonymous contributors, remains one of the most reliable
sources of information on the Internet. Wikipedia's sixty two million
articles are edited by over one hundred thousand contributors, and
these contributors have somehow done a better job maintaining the
quality and validity of the information than any of the
(00:45):
information provided by any of the major platforms today. Wikipedia
is funded entirely through the Wikimedia Foundation, a nonprofit organization
funded through public donations and grants, and has the staff
of over seven hundred people. Yeah, and I'll thank you.
Most people have absolutely no idea how intricate the webs
encyclopedia is. And today I'm joined by Gorilla Warfare, a
(01:07):
Wikipedia editor that's made over one hundred and thirty thousand
edits in the span of eighteen years across two accounts.
You'll likely know her better as researcher and critic Molly White,
the creator of the Citation Needed newsletter and web three
is going great. All right, Molly, thank you for joining me,
(01:28):
Thanks for having me. So, I don't know how I
was to ask this book, Why should people actually care
about Wikipedia today?
Speaker 2 (01:37):
Well, I mean, I think Wikipedia is one of the
highest quality resources there is on the web today. I
think everyone uses it sometimes without even realizing it. Given
that Wikipedia data powers so much of you know, AI chat,
GPT responses or home assistants that are answering your questions
(01:58):
or sirie. So I think keeping the quality of information
on Wikipedia high, or even improving it even further, is
one of the most important things that we need to
be focusing on.
Speaker 1 (02:10):
But one of the big critiques of it. And I know,
and this is very much a podcast for people who
might be techy, but people also who are just being
exposed to tech outside of the consumer realm. How is
it reliable? Because that is the big that's the big question.
How is this something you can trust?
Speaker 2 (02:27):
Well, I think, like a lot of things, you need
to take it with a grain of salt and understand
that there is variation on Wikipedia between you know, some
very high quality articles that have been reviewed by hundreds
of people versus some that are fairly new and you know,
being created still. So I think it's important to realize
that there is the possibility that whatever you're reading is
(02:49):
not reliable, and there is some onus on the reader
to verify that what they're reading is correct. But I
think that generally speaking, the quality control and Wikipedia is
actually pretty good. There are a lot of editors who
are constantly maintaining the platform, making sure that the material
(03:09):
on there is you know, well sourced, it's coming from
high quality, reputable publications, and that you know, it is
meeting the criteria that Wikipedia puts in place. And I
think the result of that is that the content on
Wikipedia tends to be very high quality.
Speaker 1 (03:27):
And how does the actual moderation work, So what goes
into a Wikipedia page?
Speaker 2 (03:33):
Well, it's a little bit ad hoc. There's not really
a process in which every page has to go through
a set of criteria or anything like that. Generally speaking,
anyone can edit Wikipedia and you know, contribute to the
best of their ability, and when someone does something that
(03:54):
is not in line with Wikipedia's requirements, then hopefully, you know,
the idea is that hopefully someone will come along and
notice that, revert the change or improve it. So that
it does meet the criteria. And I realize that sounds
very like slap dash, and it is to some extent,
but because of the sort of processes that have developed
(04:15):
over the years, it actually works pretty well. Where most
pages do you get you know, a fairly strong set
of people who are you know, taking a look at
the changes that are coming in, making sure that they
are appropriate for the encyclopedia and you know, allowing them
to remain or you know, discussing them and reverting them
(04:36):
if necessary.
Speaker 1 (04:38):
Talk to me a little bit about that process though.
Speaker 2 (04:40):
What happens well someone you know, if someone goes and
makes an edit to a page. Generally speaking, there are
you know a fair number of people who are patrolling
recent changes to the encyclopedia, you know, not necessarily even
watching that page specifically, but just looking at new edits
that are coming in and you know, checking them for reliability.
(05:01):
You know, is there a source included, does the edit
past this, you know, the smell check. There's also some
automated processes involved that will try to filter out edits
that are abusive or you know, match certain patterns. And
then there's the you know, there's a lot of editors
who have you know, various pages that they're interested in
(05:24):
or they have expertise in on what's called their watch list,
and so they take a look at changes to those
pages every once in a while and see if you know,
everything looks all right or if there needs to be
improvements or edits to you know, what's been added to
bring them back into line with the quality that we expect.
Speaker 1 (05:42):
So is there an organization? I know there's the Wikimedia Foundation,
we know that, but is there is there a moderate
to chat? Is there a place where people congregate or
is this just entirely I don't want to say decentralized,
but disorganized.
Speaker 2 (05:58):
Well, it's certainly disorganized there. Yeah, so there are there
are actually kind of a bunch of places where people congregate,
and some people use some of them and some people don't.
But you know, there's there are various places on the
encyclopedia that are you know, project specific pages where people
(06:19):
discuss issues that are coming up or flag you know,
things that need more attention. You know, not everyone has
like the ability, for example, to block an editor who's
being disruptive, So editors who can't do that themselves will
report it for people who can and then there's you know,
more real time places like we still use IRC, believe
(06:40):
it or not, but there's also.
Speaker 1 (06:42):
For that and for listeners. By the way, IRC is
what a thirty year old sharewab product.
Speaker 2 (06:48):
Yeah, it's basically one of the first online chat protocols.
Speaker 1 (06:53):
I was a polarist IRC guy. But for listeners as well,
discord is very very much I believe Discorde is actually
somehow built on IOCA. That's a different episode thought.
Speaker 2 (07:03):
I'm not sure if it actually is or if it
just interoperates well with it. But there's also a Wikipedia
Discord now for people who don't want to go figure
out how IRC works. So you know, there's a bunch
of different places where people can chat and talk about
working on the project, different formats.
Speaker 1 (07:22):
And there's no compensation of any kind.
Speaker 2 (07:25):
Correct, not unless you are, you know, an employee of
the Wikimedia Foundation, who is you know, writing the actual
media wiki software or you know, performing one of those tasks.
But that's a very small organization in comparison to the
number of people who edit the encyclopedia, the vast majority
of whom do so for you know, for free. Technically,
(07:45):
you there are some people who edit Wikipedia articles for
pay as a part of like PR strategies and stuff.
But that's a whole can of worms and it's somewhat controversial.
Speaker 1 (07:56):
Well, that's actually a good question. So I run a
p off and I know that that people will come
along on both sides and say, can you get me
a Wikipedia page? And then others will say for this
incredibly large sum of money, I can do this for you.
And it feels like those people are fighting a losing battle.
It's one of the rare cases where capitalism can't really win.
(08:18):
And how is it that Wikipedia is so resistant to
that kind of stuff?
Speaker 2 (08:23):
Right, So there's a lot of resistance from the wikimedia
community towards people who are editing for promotional purposes, because,
you know, the whole point of the project is it
supposed to be an encyclopedia, not an advertising space, not
a resume, you know, not a place to promote your
business or your product. But of course people want information
(08:46):
on Wikipedia about them, and so there are some people
who are able to sort of tread that line where
they understand wikimedia policies very well. They understand, you know,
what is allowed as far as writing about, you know,
a company without being promotional they understand what that you know,
organization or company or person would need to accomplish in
(09:08):
order to you know, achieve those notability requirements, and then
they can write about them, and they do so very transparently.
You know, they disclose that they are doing so for compensation,
you know that they've been hired by that person or organization,
and they go through the process and it's very you know,
carefully orchestrated. But there are also a bunch of people
(09:29):
who do so sort of on the you know, on
the sly and they don't really disclose that they're doing
it for pay. Those are often the people who will
sort of cold call you, like if you've gotten emails,
they're like, hey, I can make you Wikipedia page. A
lot of those are sort of scammy, uh, not so
reputable organizations, and you know, generally speaking, they actually have
(09:52):
a pretty hard time doing what they claim that they
can do because they don't tend to understand the policy
as well. They try to shoe horn articles into Wikipedia
about subjects that are not notable or that are too promotional,
and so they tend to get taken down, much to
the dismay of the people who pay them. Quite a
lot of money. So if you're listening to this and
(10:12):
you've heard from one of those people, I would not
recommend hiring them.
Speaker 1 (10:16):
Just become more important, which I guess is kind of
the question. How does Wikipedia actually judge notoriety? What is important?
Speaker 2 (10:24):
Yeah, So, generally speaking, the notability requirements is, you know,
it basically goes on how much coverage a subject has
received in reliable sources that are independent of the subject.
So you know, if a big newspaper or several newspapers
write about you in detail, chances are you might be notable,
(10:48):
en not for Wikipedia article. On the flip side, if
you're just publishing press releases that's not independent of you,
that probably doesn't qualify. If there's just one brief mention
of you a reliable source someplays, that's probably not sufficient.
So that's sort of the general guideline. There are other
sort of more specific requirements for some sorts of like
(11:11):
specific types of topics, you know, like sports players and
things like that. But generally speaking, even if someone meets
one of those guidelines, they've already met that general guideline,
where you know, it's just about the amount of coverage
in reliable sources and that can be pressed, that can
be academia, can be books, you know, anything that sort
of qualifies as a reliable source.
Speaker 1 (11:46):
Now recently seen it, one of the largest tech sites
actually found itself downgraded in reliability or were you aware
of that situation? Did you see that happen?
Speaker 2 (11:57):
I saw the conversations about it. I wasn't a participant
in them, but I sort of watched it happen.
Speaker 1 (12:02):
Yeah, So for the listeners seen it has hundreds of
that hundreds of millions of views a month. It's one
of the most notable websites, frankly, and has been for decades.
So what happened there? Why was this significant website considered
less reliable by Wikipedia?
Speaker 2 (12:22):
Yeah? So Wikipedia has these discussions pretty frequently about different
sources and whether or not those sources are generally reliable
or sometimes reliable or usually not reliable. And in order
to sort of prevent people from having to have that
same conversation every single time they want to cite a
source that's heavily cited, you know, like CE or you know,
(12:44):
the New York Times or something like that, we have
this list of very commonly discussed sources where we just discuss,
you know, we outline the general reliability of that source.
And what you're referring to is a recent discussion where
some editors decided we really needed to revisit the general
reliability of c net, which was previously considered fairly reliable
(13:06):
for sourcing on Wikipedia, and the reason for that is
that there had been a pretty noticeable change in the
quality of the UH material that they were publishing, where it,
you know, it no longer had the level of accuracy
that it once did. You know, the articles that were
being published didn't seem to be being edited well or
(13:27):
fact checked well. And so we decided that, you know,
if you're just if you're planning to use c net
as a source, you should really consider this, and you know,
consider that it's probably not even as reliable as it
once was, and maybe use something else, because you know,
they've started using AI generated content and stuff like that
(13:49):
that has you know, noticeably affected the quality of the
articles that they publish.
Speaker 1 (13:55):
And what is the what are the ramifications of that downgrading?
What does that mean? Practically?
Speaker 2 (14:00):
It means that, for you know, if you're writing a
new article in Wikipedia and you want to use c
net as a source, you are somewhat discouraged from doing so. Now,
it means that where c net has been used as
a source already, editors are going to be looking to
you know, improve that sourcing, generally by swapping it out
(14:21):
with something that's more reliable. Although, you know, as notability
changes in these publications, sometimes older content that was written
or published prior to a change at the organization will
be considered reliable, whereas newer content that's being created now
and might be using you know, AI or other tools,
(14:43):
is not considered reliable anymore. So, for example, we have
a couple of sources where we'll say that, you know,
anything they published before twenty fifteen is fine, but anything
after that you should take with a grain of salt.
So I think that's partly the case with c net
now as well.
Speaker 1 (15:00):
Why do you think that Wikipedia is so much better
at quality control than I don't know, Google, Well.
Speaker 2 (15:08):
I think there's a very different set of incentives at play.
You know, Google is a profit driven corporation. They have
other motivations besides providing the person who is using their
search engine with the most relevant results. You know, that
is ostensibly what they do, but they also have motivations
(15:29):
around generating ad revenue and clicks for you know, different
groups that are using their service or advertising with them.
Wikipedia doesn't have those sometimes conflicting incentives. You know, the
goal of Wikipedia is to deliver the highest quality information
that we can, and we're not you know, there's no advertisements.
(15:51):
We're not trying to surface some specific set of content
for people based on who is paying, you know, or
or not paying. It's really just that one incentive, and
I think that simplifies things because we can all sort
of get behind that one goal and you know, we
aren't trying to juggle multiple things that are sometimes very
much in conflict with one another, as I think we've
(16:12):
seen with Google, where you know, that goal of providing
the highest quality information to the search user has been
very much challenged by the other incentives at play.
Speaker 1 (16:25):
So you've edited Wikipedia for about eighteen years. I think
the simplest question is why do you still do it?
Why did you start? Why do you continue? Well?
Speaker 2 (16:35):
I started because I was a curious kid who discovered
that I could and that was exciting for me, and
I have I don't know. I think there are just
some people out there who have like a very specific
type of brain where Wikipedia just like sort of tickles
it just right, you know, and like doing that kind
of editing and you know, curation just appeals to me.
(16:58):
I find it really enjoyable. But I also think that,
you know, the project is a really wonderful, uh creation,
you know, and I think it provides a really important
service to the people who have access to it, which
is almost anyone. I mean, it's anyone with an Internet connection,
and there are even ways to get access to Wikipedia
(17:19):
without an Internet connection. And so I think that, you know,
maintaining a quality source of high you know, reliability material
is incredibly important and arguably getting only more important as
other sources of that same type of content, you know,
are becoming degraded in the way that you just described
with Google for example. So you know, I'm very passionate
(17:40):
about it. I think providing high quality information to everyone
is one of the most important things that sort of
humanity needs to achieve, and so I find that very
strongly motivating.
Speaker 1 (17:53):
And is there any pressure a toll from the Wikimedia Foundation?
Do they push people around a toll A not going
either way. I'm just wondering what influence they have on
the platform.
Speaker 2 (18:05):
Yeah, it's a kind of an unusual relationship between the
Wi Community Foundation and the editing community. It's sometimes a
little bit adversarial, which is sometimes a little confusing to
people who are new to the project. But I would
say no, the Wi Commedia Foundation generally takes a pretty
hands off approach towards the projects that you know, it
(18:26):
kind of supports. The Wikimedia Foundation almost never comes in
and says, hey, you need to delete that, or you
need to you know, do something very specific with the content.
There are some very edge cases with like legal requests
and things like that where they can sometimes do that,
but they are very very conservative on when they will
(18:47):
do that, and for the most part they take that
very hands off approach. The Wi Community Foundation is mostly
concerned with, you know, paying the bills, keeping the site online,
developing the soft where that actually powers the encyclopedia and
the editing infrastructure and all that, and then you know,
trying to contribute and grow the editing community and that
(19:08):
type of thing. But they are pretty hands off when
it comes to editing, which I think is actually quite good.
It allows the community to you know, organically develop its
own policies and protocols and things like that, and you know,
without the Wikimedia Foundation sort of putting its thumb on
the scale.
Speaker 1 (19:29):
So more practical question, as an editor, what do you do?
Is it just editing? The reason I ask this is
you've provided me with your Wikipedia editor page, and there's
things like on block reblog, blog thanks. What do you
do as an editor? What is what are these numbers about?
Because it feels like there's kind of a niche social
network and inside it.
Speaker 2 (19:51):
Yeah, there's a lot that happens behind the scenes besides
just you know, opening up a page and writing something new.
I do do so of that, you know, I do
write pages from scratch once in a while, or I'll
go in and edit something if a page is missing
something or need something. But there's also a whole bunch
of other sort of administrative work I guess that goes
(20:12):
on behind the scenes, where I will sometimes you know,
patrol articles. That's sort of what I had described earlier,
where people will watch the recent changes to the project
and try to just filter out anything that looks abusive
or you know, disruptive or honestly just less than productive.
Sometimes I try to, you know, go around and welcome
(20:33):
new editors who are just joining the projects and hims
they need help, and so I'll try to sort of,
you know, show them the ropes a little bit. And
then there's you know, the blocking and deletion and things
like that, where people who are known as administrators on
the project can block an editor who is regularly being
disruptive or delete a page if it doesn't meet the
(20:56):
requirements that we have defined in our policies and things
like that, and so I do some of that as well.
Speaker 1 (21:02):
So how does one become an administrator?
Speaker 2 (21:06):
So there's a whole process called request for admanship where
you either put yourself up, you nominate yourself as a candidate,
or someone else who thinks that you would be a
good administrator will do so. And then there's a long
process where you answer some questions and then the wikimedia
community votes pretty much on whether or not you you know,
(21:27):
have the experience, the you know, the mindset basically, the
right attitude towards contributing, and then there's a vote pretty
much and if you pass the threshold, then you are
nominated or you know, you become an administrator at that point.
So it's something that happens, you know, fairly regularly. I
(21:48):
did it in twenty ten, I think but you know,
various people go through it pretty often.
Speaker 1 (22:06):
Have you seen more people joining as editors or is
there more people or less people editing these days?
Speaker 2 (22:15):
I would say there's been sort of a slow decline
in the number of people who are joining if you
look at you know, people who join as like a
user and then continue to edit somewhat frequently versus you know, not.
I'm not just talking about people who like correct a
typo once in a while, but the sort of like
regular contributors is sort of yeah, it's it's either flatlining
(22:41):
or declining to some extent. Uh, And it has been
for a really long time, and so it's sort of
this continual discussion that happens within Wikipedia about you know what,
what do we do about that? How do we encourage
more people to join, How do we make the project
more welcoming to new people who want to get involved?
You know, something that I try to focus on as well,
(23:02):
just because I think it's really important to not only
contribute to the encyclopedia, but make sure that other people
are aware that they can and you know, doing so
if they have the desire to do so.
Speaker 1 (23:12):
Are they training materials? Is there an onboarding process?
Speaker 2 (23:16):
There are. In fact, there are many of such things
that different people have created, and the Wi Community Foundation
themselves has gotten a little more involved in recent years
and trying to make that sign up process a little
bit more friendly, so you know, when you first create
an account, it doesn't just drop you into the editor
like good luck, you know, which is kind of what
it used to do. Now there's some little widgets and
(23:40):
things that will guide you through making your first edit
and you know, finding an article that maybe needs some
improvement so you can give it a shot. Those types
of things are being developed, and then there's you know,
community resources to try to encourage people to do the same,
which are just created by different editors for various purposes.
Speaker 1 (24:00):
Is there any automation Yep, there's there's quite a lot.
Speaker 2 (24:05):
There are Wikipedia editing bots that uh perform various tasks,
everything from you know, fixing vandalism to you know, introducing
archive links so that if a source link goes down,
you know you can still get a copy of that
source material, all kinds of things like that. There's anti
spam stuff, you know, where it looks for spammy links
(24:27):
and removes it. So yeah, there's there's quite a bit
of automation, both by community members. Well, so some of
it's yeah, some of it's community members who who you know,
create bots and maintain them. I would say it's probably
the majority community members, but there you know, there's also
the media Wiki software itself, which is maintained by the
Wikimedia Foundation, but they don't really do much of the
(24:50):
editing side of things, so most of the editing stuff
is community based.
Speaker 1 (24:54):
Has there been any discussion of AI or integrating AI
or anything like that.
Speaker 2 (25:00):
There has been quite a lot of discussion of AI,
and it's mostly been around two things. The sourcing situation.
So can you consider AI generated material to be a
reliable source? Generally speaking? The answer has been, know that
you have to be very cautious when you know, typing
(25:21):
something into chat gpt, because a lot of times you'll
get back something that is not accurate at all. And
then there's also been discussion around can you use AI
generated content, you know, as you know, can you tell
chat gpt or something like that to go make an
edit for you know, can you use content that chat
gpt has written, as you know, content that you put
(25:42):
into a Wikipedia article, And the answer there has largely
been yeah, you can, but you have to be really
careful about it, and you have to sort of take
full responsibility for anything that you have generated and put
into a Wikipedia article, because you know, it's still up
to you to fact check it and things like that.
(26:03):
As far as integrating AI into Wikipedia to try to
i don't know, generate article content or something more generally,
that is not something that the community has seemed particularly
interested in doing, and I suspect it would introduce a
lot of problems. There were some sort of past attempts
at doing things sort of like that in the translation space,
(26:27):
where the wikimedia software would try to encourage people to
translate articles from one version of Wikipedia to another, and
that was a largely unsuccessful experiment, I think, where too
many people were just pasting in machine translation or you know,
not doing the right kinds of quality control. So we
have to be really cautious around that type of thing.
Speaker 1 (26:50):
So I think it was a year or two ago
Elon Musk said he wanted to buy Wikipedia for a
billion dollars. Just so well clear that is impossible.
Speaker 2 (27:00):
Right, yeah, I mean Wikipedia is not for sale. I
don't think there's any you know, genuine interest on either
side for Wikipedia to be sold. I think Elon is
also particularly not interested in it. He just has a
long standing beef with Wikipedia.
Speaker 1 (27:17):
Which understandably because they tell the truth, which is not
a big thing for mister Musk. Now, so you've done
a lot of advocacy recently about editing and saying why
people should become editors, But I kind of want to
hear your sales. Why should the average person edit Wikipedia?
Speaker 2 (27:35):
Well, I mean, I think, honestly, there's probably one hundred
answers for that, and it depends very much on the person.
You know, I do it because I love it. I
find it really enjoyable. Some people do it because they
think that the resource of Wikipedia is incredibly important and
needs to be maintained. You know, that's also a factor
for me. But I think it's okay to also just
(27:57):
do it just for fun. You know, I think you
don't have to have some big driving motivation behind it.
But I do think that, you know, especially now as
the web is somewhat under threat from this sort of
AI generated junk that's being you know, becoming so prevalent
in search results in any particular website that you go visit,
(28:21):
it's really important to make sure that there is this
you know, human reviewed material out there that is as
high quality as possible and that really values you know,
the quality of information, accuracy, reliability, you know, neutral point
of view, those types of things above just you know,
spitting out as much content as possible with little regard
(28:44):
for its accuracy or reliability. And so I think that
you know, now more than ever, projects like Wikipedia are
incredibly important, and you know, maintaining them as other parts
of the web begin to degrade, is going to only
be become, you know, a more important thing to do.
Speaker 1 (29:03):
It kind of feels like as this prevalence of AI
generated content grows, we kind of need user generated content well,
never as it's being killed off.
Speaker 2 (29:15):
Yeah, it's kind of ironic. I think, you know that
we're starting to see this proliferation of this AI generated
content that's really just quantity over quality, and in doing so,
it's sort of killing the quality content that is often
the source material for the AI itself. You know, we're
seeing this decline in journalism, for example, where some outlets
(29:38):
are laying off their media teams, you know, instead of
instead hoping that they can use AI to just churn
out articles. But you know those same AI tools that
they're using were trained on journalism. They were trained on
the types of stuff that is no longer being created,
and so you end up with this potentially really circular
situation where AI could start ingesting AI generated content and
(30:03):
entering this really massive quality spiral.
Speaker 1 (30:05):
I think what Jason Sadowski calls habsburg AI. Yeah, the
inbreeding about it, keep going, sorry.
Speaker 2 (30:12):
Yeah, no, it's a great terms. He always has great
terms for things. But yeah, I mean I think that
you know, we really people are very excited about AI,
and they believe that AI will just continue to get
better and better and better without really thinking that critically
about what is required to create like a high quality,
(30:33):
you know, large language model, and something like Wikipedia is honestly,
like incredibly important to creating any quality large language model.
I mean, I think every large language model out there
pretty much uses Wikipedia as a source material.
Speaker 1 (30:47):
Yeah, Chat GPT was trained on it, right.
Speaker 2 (30:49):
Yeah, I mean I think basically all of them were.
And so you know, if you enter that very circular
loop where that training material gets worse and worse and worse,
I think AI will get worse and worse and worse.
And so making sure that there is you know, very
high quality information out there that is not being in shitified.
(31:10):
To use Corey Doctor's term, bye by this AI, you know,
junk that's being created is I think very important just
to you know, general knowledge, but also I think it
should be very important to those people who care so
much about AI.
Speaker 1 (31:28):
Yeah. It's ironic as well because the sources of information
we've come to rely upon, like Google and Being to
some extent I guess, are so dependent on using generator comment.
But really it just they've choked it. They're choking it
as we speak. Is Wikipedia seeing less traffic or more traffic?
Is it? How is this affecting it?
Speaker 2 (31:49):
I honestly don't know. I guess I could look at
Wikipedia traffic that's my dub. Yeah, But I mean I
think there is, you know, to some extent, with Google
and these various software projects that ingest Wikipedia data. You know,
sometimes they do draw that attention away from the source,
(32:10):
you know, where they'll just highlight the first paragraph of
Wikipedia and search results and people never click through to
the Wikipedia article. That's a pretty common phenomenon. You know.
News sources are sometimes upset about that also. But I
think one thing that's useful about Wikipedia is that Wikipedia
doesn't really have the same incentive to draw clicks as
(32:30):
you know, an ad supported news source might, and so.
Speaker 1 (32:35):
To selected on search.
Speaker 2 (32:38):
But it is, yeah, exactly, It's like to some extent,
as long as the information is getting out there, we
don't really care how and so you know, I guess
there's some question of like, is it actually bad for
traffic to be redirected away from Wikipedia? I could probably
make that argument in both directions.
Speaker 1 (32:57):
Well, Molly, thank you so much for joining me today.
Where can people find you?
Speaker 2 (33:02):
You can find me at Mollywaite dot net. I also
read a newsletter at Citation Needed dot news.
Speaker 1 (33:09):
Well, thank you so much for joining me.
Speaker 2 (33:11):
Thanks for having me.
Speaker 1 (33:13):
Now, listeners, I'm going to do something a little bit
off based, if you will. Molly and I just talked
about Wikipedia, and I realize it feels weird to advocate
for something on a objective media platform, though I think
we all agree that I have my biases and my
things that I care about more and that I'm angry about,
and I think we all do. But I must state
(33:34):
how important keeping Wikipedia alive is. This platform, despite the
fact that's editable by everyone, is more reliable than Google
is at the moment, it's more reliable than a lot
of media publications. As we speak, the user generated Internet
is being destroyed, it is being sold off for parts,
it's being turned into a rot economic catastrophe. Something like
(33:57):
Wikipedia is truly important. Really. Again, it's weird to advocate
for any product or thing, but I encourage you, and
I know it's weird to say, but please go and
edit Wikipedia. Please sit down and edit whatever it is.
Learn the ropes. This is one of the few community
organized Internet things that actually exist that we can help with.
It's a thing you can do today to change the Internet,
(34:19):
to fight for what's right. And yeah, these publications are
funded in part by tech publications. The Wikimedia Foundation has
taken money from Google. But as you've heard from Molly,
who I trust deeply and you should do, they do
have a firewall between them. This is one way you
can fight back, and I implore you to do so,
(34:41):
even if it's one editor, even if it's one page.
You keep an eye on, follow the training documents, join
the party. Please try. I understand that it's impossible to
ask for money, and I should never do so, but
your time, your attention on Wikipedia is genuinely important. It
sounds silly. I know we've all kind of thought it's
(35:02):
just Wikipedia, anyone can edit it. But I'm worried. I'm
worried for the Internet as it stands. I'm worried the
Internet sources are going to become more centralized, more focused
on the big take platforms. Please protect what's left of
the good Internet. Edit Wikipedia today. I'm serious. At times,
(35:22):
it can feel a little hopeless out there. It can
feel like there's nothing we can do against these trillion
dollar enterprises, and to some extent, there's nothing we can do.
We really can't. We can't stop Sundar Pashai, we can't
stop Sam Ortman. What we can do is help reinforce
what made the Internet great. What we can do is
(35:43):
contribute to open source. What we can do is edit Wikipedia,
even a little, Even those little contributions matter. This is
what made the Internet what it is today, and we
can fight for it and we can protect it. Thank
you for listening, Thank you for listening to Better Offline.
(36:10):
The editor and composer of the Better Offline theme song
is Matasowski. You can check out more of his music
and audio projects at Matasowski dot com. M A T
T O S O W s KI dot com. You
can email me at easy at better offline dot com
or check out better offline dot com to find my
newsletter and more links to this podcast. Thank you so
(36:31):
much for listening. Better Offline is a production of cool
Zone Media. For more from cool Zone Media, visit our
website cool Zonemedia dot com, or check us out on
the iHeartRadio
Speaker 2 (36:41):
App, Apple Podcasts, or wherever you get your podcasts.
All Zone Media. Hello and welcome to Better Offline. I'm
your host ed zetron. Wikipedia is a free online encyclopedia
used by over a billion people a month, one that,
(00:24):
despite being maintained entirely by an army of free and
at times anonymous contributors, remains one of the most reliable
sources of information on the Internet. Wikipedia's sixty two million
articles are edited by over one hundred thousand contributors, and
these contributors have somehow done a better job maintaining the
quality and validity of the information than any of the
(00:45):
information provided by any of the major platforms today. Wikipedia
is funded entirely through the Wikimedia Foundation, a nonprofit organization
funded through public donations and grants, and has the staff
of over seven hundred people. Yeah, and I'll thank you.
Most people have absolutely no idea how intricate the webs
encyclopedia is. And today I'm joined by Gorilla Warfare, a
(01:07):
Wikipedia editor that's made over one hundred and thirty thousand
edits in the span of eighteen years across two accounts.
You'll likely know her better as researcher and critic Molly White,
the creator of the Citation Needed newsletter and web three
is going great. All right, Molly, thank you for joining me,
(01:28):
Thanks for having me. So, I don't know how I
was to ask this book, Why should people actually care
about Wikipedia today?
Speaker 2 (01:37):
Well, I mean, I think Wikipedia is one of the
highest quality resources there is on the web today. I
think everyone uses it sometimes without even realizing it. Given
that Wikipedia data powers so much of you know, AI chat,
GPT responses or home assistants that are answering your questions
(01:58):
or sirie. So I think keeping the quality of information
on Wikipedia high, or even improving it even further, is
one of the most important things that we need to
be focusing on.
Speaker 1 (02:10):
But one of the big critiques of it. And I know,
and this is very much a podcast for people who
might be techy, but people also who are just being
exposed to tech outside of the consumer realm. How is
it reliable? Because that is the big that's the big question.
How is this something you can trust?
Speaker 2 (02:27):
Well, I think, like a lot of things, you need
to take it with a grain of salt and understand
that there is variation on Wikipedia between you know, some
very high quality articles that have been reviewed by hundreds
of people versus some that are fairly new and you know,
being created still. So I think it's important to realize
that there is the possibility that whatever you're reading is
(02:49):
not reliable, and there is some onus on the reader
to verify that what they're reading is correct. But I
think that generally speaking, the quality control and Wikipedia is
actually pretty good. There are a lot of editors who
are constantly maintaining the platform, making sure that the material
(03:09):
on there is you know, well sourced, it's coming from
high quality, reputable publications, and that you know, it is
meeting the criteria that Wikipedia puts in place. And I
think the result of that is that the content on
Wikipedia tends to be very high quality.
Speaker 1 (03:27):
And how does the actual moderation work, So what goes
into a Wikipedia page?
Speaker 2 (03:33):
Well, it's a little bit ad hoc. There's not really
a process in which every page has to go through
a set of criteria or anything like that. Generally speaking,
anyone can edit Wikipedia and you know, contribute to the
best of their ability, and when someone does something that
(03:54):
is not in line with Wikipedia's requirements, then hopefully, you know,
the idea is that hopefully someone will come along and
notice that, revert the change or improve it. So that
it does meet the criteria. And I realize that sounds
very like slap dash, and it is to some extent,
but because of the sort of processes that have developed
(04:15):
over the years, it actually works pretty well. Where most
pages do you get you know, a fairly strong set
of people who are you know, taking a look at
the changes that are coming in, making sure that they
are appropriate for the encyclopedia and you know, allowing them
to remain or you know, discussing them and reverting them
(04:36):
if necessary.
Speaker 1 (04:38):
Talk to me a little bit about that process though.
Speaker 2 (04:40):
What happens well someone you know, if someone goes and
makes an edit to a page. Generally speaking, there are
you know a fair number of people who are patrolling
recent changes to the encyclopedia, you know, not necessarily even
watching that page specifically, but just looking at new edits
that are coming in and you know, checking them for reliability.
(05:01):
You know, is there a source included, does the edit
past this, you know, the smell check. There's also some
automated processes involved that will try to filter out edits
that are abusive or you know, match certain patterns. And
then there's the you know, there's a lot of editors
who have you know, various pages that they're interested in
(05:24):
or they have expertise in on what's called their watch list,
and so they take a look at changes to those
pages every once in a while and see if you know,
everything looks all right or if there needs to be
improvements or edits to you know, what's been added to
bring them back into line with the quality that we expect.
Speaker 1 (05:42):
So is there an organization? I know there's the Wikimedia Foundation,
we know that, but is there is there a moderate
to chat? Is there a place where people congregate or
is this just entirely I don't want to say decentralized,
but disorganized.
Speaker 2 (05:58):
Well, it's certainly disorganized there. Yeah, so there are there
are actually kind of a bunch of places where people congregate,
and some people use some of them and some people don't.
But you know, there's there are various places on the
encyclopedia that are you know, project specific pages where people
(06:19):
discuss issues that are coming up or flag you know,
things that need more attention. You know, not everyone has
like the ability, for example, to block an editor who's
being disruptive, So editors who can't do that themselves will
report it for people who can and then there's you know,
more real time places like we still use IRC, believe
(06:40):
it or not, but there's also.
Speaker 1 (06:42):
For that and for listeners. By the way, IRC is
what a thirty year old sharewab product.
Speaker 2 (06:48):
Yeah, it's basically one of the first online chat protocols.
Speaker 1 (06:53):
I was a polarist IRC guy. But for listeners as well,
discord is very very much I believe Discorde is actually
somehow built on IOCA. That's a different episode thought.
Speaker 2 (07:03):
I'm not sure if it actually is or if it
just interoperates well with it. But there's also a Wikipedia
Discord now for people who don't want to go figure
out how IRC works. So you know, there's a bunch
of different places where people can chat and talk about
working on the project, different formats.
Speaker 1 (07:22):
And there's no compensation of any kind.
Speaker 2 (07:25):
Correct, not unless you are, you know, an employee of
the Wikimedia Foundation, who is you know, writing the actual
media wiki software or you know, performing one of those tasks.
But that's a very small organization in comparison to the
number of people who edit the encyclopedia, the vast majority
of whom do so for you know, for free. Technically,
(07:45):
you there are some people who edit Wikipedia articles for
pay as a part of like PR strategies and stuff.
But that's a whole can of worms and it's somewhat controversial.
Speaker 1 (07:56):
Well, that's actually a good question. So I run a
p off and I know that that people will come
along on both sides and say, can you get me
a Wikipedia page? And then others will say for this
incredibly large sum of money, I can do this for you.
And it feels like those people are fighting a losing battle.
It's one of the rare cases where capitalism can't really win.
(08:18):
And how is it that Wikipedia is so resistant to
that kind of stuff?
Speaker 2 (08:23):
Right, So there's a lot of resistance from the wikimedia
community towards people who are editing for promotional purposes, because,
you know, the whole point of the project is it
supposed to be an encyclopedia, not an advertising space, not
a resume, you know, not a place to promote your
business or your product. But of course people want information
(08:46):
on Wikipedia about them, and so there are some people
who are able to sort of tread that line where
they understand wikimedia policies very well. They understand, you know,
what is allowed as far as writing about, you know,
a company without being promotional they understand what that you know,
organization or company or person would need to accomplish in
(09:08):
order to you know, achieve those notability requirements, and then
they can write about them, and they do so very transparently.
You know, they disclose that they are doing so for compensation,
you know that they've been hired by that person or organization,
and they go through the process and it's very you know,
carefully orchestrated. But there are also a bunch of people
(09:29):
who do so sort of on the you know, on
the sly and they don't really disclose that they're doing
it for pay. Those are often the people who will
sort of cold call you, like if you've gotten emails,
they're like, hey, I can make you Wikipedia page. A
lot of those are sort of scammy, uh, not so
reputable organizations, and you know, generally speaking, they actually have
(09:52):
a pretty hard time doing what they claim that they
can do because they don't tend to understand the policy
as well. They try to shoe horn articles into Wikipedia
about subjects that are not notable or that are too promotional,
and so they tend to get taken down, much to
the dismay of the people who pay them. Quite a
lot of money. So if you're listening to this and
(10:12):
you've heard from one of those people, I would not
recommend hiring them.
Speaker 1 (10:16):
Just become more important, which I guess is kind of
the question. How does Wikipedia actually judge notoriety? What is important?
Speaker 2 (10:24):
Yeah, So, generally speaking, the notability requirements is, you know,
it basically goes on how much coverage a subject has
received in reliable sources that are independent of the subject.
So you know, if a big newspaper or several newspapers
write about you in detail, chances are you might be notable,
(10:48):
en not for Wikipedia article. On the flip side, if
you're just publishing press releases that's not independent of you,
that probably doesn't qualify. If there's just one brief mention
of you a reliable source someplays, that's probably not sufficient.
So that's sort of the general guideline. There are other
sort of more specific requirements for some sorts of like
(11:11):
specific types of topics, you know, like sports players and
things like that. But generally speaking, even if someone meets
one of those guidelines, they've already met that general guideline,
where you know, it's just about the amount of coverage
in reliable sources and that can be pressed, that can
be academia, can be books, you know, anything that sort
of qualifies as a reliable source.
Speaker 1 (11:46):
Now recently seen it, one of the largest tech sites
actually found itself downgraded in reliability or were you aware
of that situation? Did you see that happen?
Speaker 2 (11:57):
I saw the conversations about it. I wasn't a participant
in them, but I sort of watched it happen.
Speaker 1 (12:02):
Yeah, So for the listeners seen it has hundreds of
that hundreds of millions of views a month. It's one
of the most notable websites, frankly, and has been for decades.
So what happened there? Why was this significant website considered
less reliable by Wikipedia?
Speaker 2 (12:22):
Yeah? So Wikipedia has these discussions pretty frequently about different
sources and whether or not those sources are generally reliable
or sometimes reliable or usually not reliable. And in order
to sort of prevent people from having to have that
same conversation every single time they want to cite a
source that's heavily cited, you know, like CE or you know,
(12:44):
the New York Times or something like that, we have
this list of very commonly discussed sources where we just discuss,
you know, we outline the general reliability of that source.
And what you're referring to is a recent discussion where
some editors decided we really needed to revisit the general
reliability of c net, which was previously considered fairly reliable
(13:06):
for sourcing on Wikipedia, and the reason for that is
that there had been a pretty noticeable change in the
quality of the UH material that they were publishing, where it,
you know, it no longer had the level of accuracy
that it once did. You know, the articles that were
being published didn't seem to be being edited well or
(13:27):
fact checked well. And so we decided that, you know,
if you're just if you're planning to use c net
as a source, you should really consider this, and you know,
consider that it's probably not even as reliable as it
once was, and maybe use something else, because you know,
they've started using AI generated content and stuff like that
(13:49):
that has you know, noticeably affected the quality of the
articles that they publish.
Speaker 1 (13:55):
And what is the what are the ramifications of that downgrading?
What does that mean? Practically?
Speaker 2 (14:00):
It means that, for you know, if you're writing a
new article in Wikipedia and you want to use c
net as a source, you are somewhat discouraged from doing so. Now,
it means that where c net has been used as
a source already, editors are going to be looking to
you know, improve that sourcing, generally by swapping it out
(14:21):
with something that's more reliable. Although, you know, as notability
changes in these publications, sometimes older content that was written
or published prior to a change at the organization will
be considered reliable, whereas newer content that's being created now
and might be using you know, AI or other tools,
(14:43):
is not considered reliable anymore. So, for example, we have
a couple of sources where we'll say that, you know,
anything they published before twenty fifteen is fine, but anything
after that you should take with a grain of salt.
So I think that's partly the case with c net
now as well.
Speaker 1 (15:00):
Why do you think that Wikipedia is so much better
at quality control than I don't know, Google, Well.
Speaker 2 (15:08):
I think there's a very different set of incentives at play.
You know, Google is a profit driven corporation. They have
other motivations besides providing the person who is using their
search engine with the most relevant results. You know, that
is ostensibly what they do, but they also have motivations
(15:29):
around generating ad revenue and clicks for you know, different
groups that are using their service or advertising with them.
Wikipedia doesn't have those sometimes conflicting incentives. You know, the
goal of Wikipedia is to deliver the highest quality information
that we can, and we're not you know, there's no advertisements.
(15:51):
We're not trying to surface some specific set of content
for people based on who is paying, you know, or
or not paying. It's really just that one incentive, and
I think that simplifies things because we can all sort
of get behind that one goal and you know, we
aren't trying to juggle multiple things that are sometimes very
much in conflict with one another, as I think we've
(16:12):
seen with Google, where you know, that goal of providing
the highest quality information to the search user has been
very much challenged by the other incentives at play.
Speaker 1 (16:25):
So you've edited Wikipedia for about eighteen years. I think
the simplest question is why do you still do it?
Why did you start? Why do you continue? Well?
Speaker 2 (16:35):
I started because I was a curious kid who discovered
that I could and that was exciting for me, and
I have I don't know. I think there are just
some people out there who have like a very specific
type of brain where Wikipedia just like sort of tickles
it just right, you know, and like doing that kind
of editing and you know, curation just appeals to me.
(16:58):
I find it really enjoyable. But I also think that,
you know, the project is a really wonderful, uh creation,
you know, and I think it provides a really important
service to the people who have access to it, which
is almost anyone. I mean, it's anyone with an Internet connection,
and there are even ways to get access to Wikipedia
(17:19):
without an Internet connection. And so I think that, you know,
maintaining a quality source of high you know, reliability material
is incredibly important and arguably getting only more important as
other sources of that same type of content, you know,
are becoming degraded in the way that you just described
with Google for example. So you know, I'm very passionate
(17:40):
about it. I think providing high quality information to everyone
is one of the most important things that sort of
humanity needs to achieve, and so I find that very
strongly motivating.
Speaker 1 (17:53):
And is there any pressure a toll from the Wikimedia Foundation?
Do they push people around a toll A not going
either way. I'm just wondering what influence they have on
the platform.
Speaker 2 (18:05):
Yeah, it's a kind of an unusual relationship between the
Wi Community Foundation and the editing community. It's sometimes a
little bit adversarial, which is sometimes a little confusing to
people who are new to the project. But I would
say no, the Wi Commedia Foundation generally takes a pretty
hands off approach towards the projects that you know, it
(18:26):
kind of supports. The Wikimedia Foundation almost never comes in
and says, hey, you need to delete that, or you
need to you know, do something very specific with the content.
There are some very edge cases with like legal requests
and things like that where they can sometimes do that,
but they are very very conservative on when they will
(18:47):
do that, and for the most part they take that
very hands off approach. The Wi Community Foundation is mostly
concerned with, you know, paying the bills, keeping the site online,
developing the soft where that actually powers the encyclopedia and
the editing infrastructure and all that, and then you know,
trying to contribute and grow the editing community and that
(19:08):
type of thing. But they are pretty hands off when
it comes to editing, which I think is actually quite good.
It allows the community to you know, organically develop its
own policies and protocols and things like that, and you know,
without the Wikimedia Foundation sort of putting its thumb on
the scale.
Speaker 1 (19:29):
So more practical question, as an editor, what do you do?
Is it just editing? The reason I ask this is
you've provided me with your Wikipedia editor page, and there's
things like on block reblog, blog thanks. What do you
do as an editor? What is what are these numbers about?
Because it feels like there's kind of a niche social
network and inside it.
Speaker 2 (19:51):
Yeah, there's a lot that happens behind the scenes besides
just you know, opening up a page and writing something new.
I do do so of that, you know, I do
write pages from scratch once in a while, or I'll
go in and edit something if a page is missing
something or need something. But there's also a whole bunch
of other sort of administrative work I guess that goes
(20:12):
on behind the scenes, where I will sometimes you know,
patrol articles. That's sort of what I had described earlier,
where people will watch the recent changes to the project
and try to just filter out anything that looks abusive
or you know, disruptive or honestly just less than productive.
Sometimes I try to, you know, go around and welcome
(20:33):
new editors who are just joining the projects and hims
they need help, and so I'll try to sort of,
you know, show them the ropes a little bit. And
then there's you know, the blocking and deletion and things
like that, where people who are known as administrators on
the project can block an editor who is regularly being
disruptive or delete a page if it doesn't meet the
(20:56):
requirements that we have defined in our policies and things
like that, and so I do some of that as well.
Speaker 1 (21:02):
So how does one become an administrator?
Speaker 2 (21:06):
So there's a whole process called request for admanship where
you either put yourself up, you nominate yourself as a candidate,
or someone else who thinks that you would be a
good administrator will do so. And then there's a long
process where you answer some questions and then the wikimedia
community votes pretty much on whether or not you you know,
(21:27):
have the experience, the you know, the mindset basically, the
right attitude towards contributing, and then there's a vote pretty
much and if you pass the threshold, then you are
nominated or you know, you become an administrator at that point.
So it's something that happens, you know, fairly regularly. I
(21:48):
did it in twenty ten, I think but you know,
various people go through it pretty often.
Speaker 1 (22:06):
Have you seen more people joining as editors or is
there more people or less people editing these days?
Speaker 2 (22:15):
I would say there's been sort of a slow decline
in the number of people who are joining if you
look at you know, people who join as like a
user and then continue to edit somewhat frequently versus you know, not.
I'm not just talking about people who like correct a
typo once in a while, but the sort of like
regular contributors is sort of yeah, it's it's either flatlining
(22:41):
or declining to some extent. Uh, And it has been
for a really long time, and so it's sort of
this continual discussion that happens within Wikipedia about you know what,
what do we do about that? How do we encourage
more people to join, How do we make the project
more welcoming to new people who want to get involved?
You know, something that I try to focus on as well,
(23:02):
just because I think it's really important to not only
contribute to the encyclopedia, but make sure that other people
are aware that they can and you know, doing so
if they have the desire to do so.
Speaker 1 (23:12):
Are they training materials? Is there an onboarding process?
Speaker 2 (23:16):
There are. In fact, there are many of such things
that different people have created, and the Wi Community Foundation
themselves has gotten a little more involved in recent years
and trying to make that sign up process a little
bit more friendly, so you know, when you first create
an account, it doesn't just drop you into the editor
like good luck, you know, which is kind of what
it used to do. Now there's some little widgets and
(23:40):
things that will guide you through making your first edit
and you know, finding an article that maybe needs some
improvement so you can give it a shot. Those types
of things are being developed, and then there's you know,
community resources to try to encourage people to do the same,
which are just created by different editors for various purposes.
Speaker 1 (24:00):
Is there any automation Yep, there's there's quite a lot.
Speaker 2 (24:05):
There are Wikipedia editing bots that uh perform various tasks,
everything from you know, fixing vandalism to you know, introducing
archive links so that if a source link goes down,
you know you can still get a copy of that
source material, all kinds of things like that. There's anti
spam stuff, you know, where it looks for spammy links
(24:27):
and removes it. So yeah, there's there's quite a bit
of automation, both by community members. Well, so some of
it's yeah, some of it's community members who who you know,
create bots and maintain them. I would say it's probably
the majority community members, but there you know, there's also
the media Wiki software itself, which is maintained by the
Wikimedia Foundation, but they don't really do much of the
(24:50):
editing side of things, so most of the editing stuff
is community based.
Speaker 1 (24:54):
Has there been any discussion of AI or integrating AI
or anything like that.
Speaker 2 (25:00):
There has been quite a lot of discussion of AI,
and it's mostly been around two things. The sourcing situation.
So can you consider AI generated material to be a
reliable source? Generally speaking? The answer has been, know that
you have to be very cautious when you know, typing
(25:21):
something into chat gpt, because a lot of times you'll
get back something that is not accurate at all. And
then there's also been discussion around can you use AI
generated content, you know, as you know, can you tell
chat gpt or something like that to go make an
edit for you know, can you use content that chat
gpt has written, as you know, content that you put
(25:42):
into a Wikipedia article, And the answer there has largely
been yeah, you can, but you have to be really
careful about it, and you have to sort of take
full responsibility for anything that you have generated and put
into a Wikipedia article, because you know, it's still up
to you to fact check it and things like that.
(26:03):
As far as integrating AI into Wikipedia to try to
i don't know, generate article content or something more generally,
that is not something that the community has seemed particularly
interested in doing, and I suspect it would introduce a
lot of problems. There were some sort of past attempts
at doing things sort of like that in the translation space,
(26:27):
where the wikimedia software would try to encourage people to
translate articles from one version of Wikipedia to another, and
that was a largely unsuccessful experiment, I think, where too
many people were just pasting in machine translation or you know,
not doing the right kinds of quality control. So we
have to be really cautious around that type of thing.
Speaker 1 (26:50):
So I think it was a year or two ago
Elon Musk said he wanted to buy Wikipedia for a
billion dollars. Just so well clear that is impossible.
Speaker 2 (27:00):
Right, yeah, I mean Wikipedia is not for sale. I
don't think there's any you know, genuine interest on either
side for Wikipedia to be sold. I think Elon is
also particularly not interested in it. He just has a
long standing beef with Wikipedia.
Speaker 1 (27:17):
Which understandably because they tell the truth, which is not
a big thing for mister Musk. Now, so you've done
a lot of advocacy recently about editing and saying why
people should become editors, But I kind of want to
hear your sales. Why should the average person edit Wikipedia?
Speaker 2 (27:35):
Well, I mean, I think, honestly, there's probably one hundred
answers for that, and it depends very much on the person.
You know, I do it because I love it. I
find it really enjoyable. Some people do it because they
think that the resource of Wikipedia is incredibly important and
needs to be maintained. You know, that's also a factor
for me. But I think it's okay to also just
(27:57):
do it just for fun. You know, I think you
don't have to have some big driving motivation behind it.
But I do think that, you know, especially now as
the web is somewhat under threat from this sort of
AI generated junk that's being you know, becoming so prevalent
in search results in any particular website that you go visit,
(28:21):
it's really important to make sure that there is this
you know, human reviewed material out there that is as
high quality as possible and that really values you know,
the quality of information, accuracy, reliability, you know, neutral point
of view, those types of things above just you know,
spitting out as much content as possible with little regard
(28:44):
for its accuracy or reliability. And so I think that
you know, now more than ever, projects like Wikipedia are
incredibly important, and you know, maintaining them as other parts
of the web begin to degrade, is going to only
be become, you know, a more important thing to do.
Speaker 1 (29:03):
It kind of feels like as this prevalence of AI
generated content grows, we kind of need user generated content well,
never as it's being killed off.
Speaker 2 (29:15):
Yeah, it's kind of ironic. I think, you know that
we're starting to see this proliferation of this AI generated
content that's really just quantity over quality, and in doing so,
it's sort of killing the quality content that is often
the source material for the AI itself. You know, we're
seeing this decline in journalism, for example, where some outlets
(29:38):
are laying off their media teams, you know, instead of
instead hoping that they can use AI to just churn
out articles. But you know those same AI tools that
they're using were trained on journalism. They were trained on
the types of stuff that is no longer being created,
and so you end up with this potentially really circular
situation where AI could start ingesting AI generated content and
(30:03):
entering this really massive quality spiral.
Speaker 1 (30:05):
I think what Jason Sadowski calls habsburg AI. Yeah, the
inbreeding about it, keep going, sorry.
Speaker 2 (30:12):
Yeah, no, it's a great terms. He always has great
terms for things. But yeah, I mean I think that
you know, we really people are very excited about AI,
and they believe that AI will just continue to get
better and better and better without really thinking that critically
about what is required to create like a high quality,
(30:33):
you know, large language model, and something like Wikipedia is honestly,
like incredibly important to creating any quality large language model.
I mean, I think every large language model out there
pretty much uses Wikipedia as a source material.
Speaker 1 (30:47):
Yeah, Chat GPT was trained on it, right.
Speaker 2 (30:49):
Yeah, I mean I think basically all of them were.
And so you know, if you enter that very circular
loop where that training material gets worse and worse and worse,
I think AI will get worse and worse and worse.
And so making sure that there is you know, very
high quality information out there that is not being in shitified.
(31:10):
To use Corey Doctor's term, bye by this AI, you know,
junk that's being created is I think very important just
to you know, general knowledge, but also I think it
should be very important to those people who care so
much about AI.
Speaker 1 (31:28):
Yeah. It's ironic as well because the sources of information
we've come to rely upon, like Google and Being to
some extent I guess, are so dependent on using generator comment.
But really it just they've choked it. They're choking it
as we speak. Is Wikipedia seeing less traffic or more traffic?
Is it? How is this affecting it?
Speaker 2 (31:49):
I honestly don't know. I guess I could look at
Wikipedia traffic that's my dub. Yeah, But I mean I
think there is, you know, to some extent, with Google
and these various software projects that ingest Wikipedia data. You know,
sometimes they do draw that attention away from the source,
(32:10):
you know, where they'll just highlight the first paragraph of
Wikipedia and search results and people never click through to
the Wikipedia article. That's a pretty common phenomenon. You know.
News sources are sometimes upset about that also. But I
think one thing that's useful about Wikipedia is that Wikipedia
doesn't really have the same incentive to draw clicks as
(32:30):
you know, an ad supported news source might, and so.
Speaker 1 (32:35):
To selected on search.
Speaker 2 (32:38):
But it is, yeah, exactly, It's like to some extent,
as long as the information is getting out there, we
don't really care how and so you know, I guess
there's some question of like, is it actually bad for
traffic to be redirected away from Wikipedia? I could probably
make that argument in both directions.
Speaker 1 (32:57):
Well, Molly, thank you so much for joining me today.
Where can people find you?
Speaker 2 (33:02):
You can find me at Mollywaite dot net. I also
read a newsletter at Citation Needed dot news.
Speaker 1 (33:09):
Well, thank you so much for joining me.
Speaker 2 (33:11):
Thanks for having me.
Speaker 1 (33:13):
Now, listeners, I'm going to do something a little bit
off based, if you will. Molly and I just talked
about Wikipedia, and I realize it feels weird to advocate
for something on a objective media platform, though I think
we all agree that I have my biases and my
things that I care about more and that I'm angry about,
and I think we all do. But I must state
(33:34):
how important keeping Wikipedia alive is. This platform, despite the
fact that's editable by everyone, is more reliable than Google
is at the moment, it's more reliable than a lot
of media publications. As we speak, the user generated Internet
is being destroyed, it is being sold off for parts,
it's being turned into a rot economic catastrophe. Something like
(33:57):
Wikipedia is truly important. Really. Again, it's weird to advocate
for any product or thing, but I encourage you, and
I know it's weird to say, but please go and
edit Wikipedia. Please sit down and edit whatever it is.
Learn the ropes. This is one of the few community
organized Internet things that actually exist that we can help with.
It's a thing you can do today to change the Internet,
(34:19):
to fight for what's right. And yeah, these publications are
funded in part by tech publications. The Wikimedia Foundation has
taken money from Google. But as you've heard from Molly,
who I trust deeply and you should do, they do
have a firewall between them. This is one way you
can fight back, and I implore you to do so,
(34:41):
even if it's one editor, even if it's one page.
You keep an eye on, follow the training documents, join
the party. Please try. I understand that it's impossible to
ask for money, and I should never do so, but
your time, your attention on Wikipedia is genuinely important. It
sounds silly. I know we've all kind of thought it's
(35:02):
just Wikipedia, anyone can edit it. But I'm worried. I'm
worried for the Internet as it stands. I'm worried the
Internet sources are going to become more centralized, more focused
on the big take platforms. Please protect what's left of
the good Internet. Edit Wikipedia today. I'm serious. At times,
(35:22):
it can feel a little hopeless out there. It can
feel like there's nothing we can do against these trillion
dollar enterprises, and to some extent, there's nothing we can do.
We really can't. We can't stop Sundar Pashai, we can't
stop Sam Ortman. What we can do is help reinforce
what made the Internet great. What we can do is
(35:43):
contribute to open source. What we can do is edit Wikipedia,
even a little, Even those little contributions matter. This is
what made the Internet what it is today, and we
can fight for it and we can protect it. Thank
you for listening, Thank you for listening to Better Offline.
(36:10):
The editor and composer of the Better Offline theme song
is Matasowski. You can check out more of his music
and audio projects at Matasowski dot com. M A T
T O S O W s KI dot com. You
can email me at easy at better offline dot com
or check out better offline dot com to find my
newsletter and more links to this podcast. Thank you so
(36:31):
much for listening. Better Offline is a production of cool
Zone Media. For more from cool Zone Media, visit our
website cool Zonemedia dot com, or check us out on
the iHeartRadio
Speaker 2 (36:41):
App, Apple Podcasts, or wherever you get your podcasts.
Popular Podcasts
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations.
The Nikki Glaser Podcast
Every week comedian and infamous roaster Nikki Glaser provides a fun, fast-paced, and brutally honest look into current pop-culture and her own personal life.
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.