At the time of posting, the website seems to be stuck on loading.
Xikipedia is a pseudo social media feed that algorithmically shows you content from Simple Wikipedia. It is made as a demonstration of how even a basic non-ML algorithm with no data from other users can quickly learn what you engage with to suggest you more similar content. No data is collected or shared here, the algorithm runs locally and the data disappears once you refresh or close the tab.
At the time of posting, the website seems to be stuck on loading.
Might be your browser, they are using ~60mb of a binary (likely WASM, ie, Web Assembly) to run the model locally, some browsers or network connections have some iffy times loading WASM blobs.
Might be your browser, they are using ~60mb of a binary (likely WASM, ie, Web Assembly) to run the model locally, some browsers or network connections have some iffy times loading WASM blobs.
Ah, yeah WASM support on mobile devices is kneecapped due to the various levels of devices, limited memory support, and that WASM support itself still isn't totally stable (or designed) even on...
Ah, yeah WASM support on mobile devices is kneecapped due to the various levels of devices, limited memory support, and that WASM support itself still isn't totally stable (or designed) even on desktop.
This is fascinating to use and see in action. I am fascinated by the negative points for every "heart" I give on some articles - some of these topics are incredible to see go up and down based on...
This is fascinating to use and see in action. I am fascinated by the negative points for every "heart" I give on some articles - some of these topics are incredible to see go up and down based on my interest. For example: Why did random articles I enjoy end up giving 3022 a science fiction film, so many negative points? In fact, just the reasoning that this system was devaluing it has made me interested - I have never seen it, likely isn't even good, but it is exactly the premise of scifi movies I am interested in - and similar to things I am interested in!
Wikipedia should really take up something like this on the official site, I am discovering so many articles that will destroy the rest of my day. I had a lot planned, now I am planning on even less because I am going down a Wikipedia hole.
I'd argue that one of the things that's kept Wikipedia out of the Enshittification whirlpool is the lack of interest in engagement. Hank Green recently posted a video going into this very topic:...
I'd argue that one of the things that's kept Wikipedia out of the Enshittification whirlpool is the lack of interest in engagement.
Wikipedia's goal is really to catalogue and explain the universe, from a human lens, for the betterment and sake of knowledge and human existence. I would say that an interface that draws more...
Wikipedia's goal is really to catalogue and explain the universe, from a human lens, for the betterment and sake of knowledge and human existence. I would say that an interface that draws more people in, to encourage research, even through a simple addon like this (they kind of do already, but through daily editorials), and they could really tap into some interesting knowledge on human interests, and inspire, create, and cultivate a wider culture to become editors (we need more of them!) and better thinkers.
We generally need more researchers in the world, and more platforms that feed the need for social feeds, with something beneficial for the mind. A bicycle of the connected mind, if you will.
Perhaps Wikipedia is the quality it currently is because the wrong kind of editor self select out of it. Maybe more engagement would dillute the quality contributions among a sea of low effort...
Perhaps Wikipedia is the quality it currently is because the wrong kind of editor self select out of it.
Maybe more engagement would dillute the quality contributions among a sea of low effort ones.
We don't really have many good things like wikipedia, we should be carefull with it.
How do you see points? Does this only value things that you open or like? Why would you like something? I see how I opened one article and it started giving me a bunch of related articles I have...
How do you see points? Does this only value things that you open or like? Why would you like something?
I see how I opened one article and it started giving me a bunch of related articles I have no interest in; how do I know if it's figuring out anything else about me?
If you are on mobile, it looks like they hide the stats box. Mine off to the right side says: should be noted, I did not 'like' or decide not to like anything about "Category:Isle of Wight". If...
If you are on mobile, it looks like they hide the stats box. Mine off to the right side says:
science: 5622
technology: 5131
human sexuality: 1140
human activities: 1076
engineering: 396
Lombardy: 325
physics: 304
chemistry: 293
electronics: 277
phenomena: 276
...
american actors by medium: -20
2020 deaths: -25
Category:Isle of Wight: -25
american television actors: -25
american television: -30
cities in the united states by state: -35
living people: -66
people: -66
given names: -1000
surnames: -1000
should be noted, I did not 'like' or decide not to like anything about "Category:Isle of Wight".
If you want to see your stats on mobile, set your webpage view to 'desktop mode' and it will zoom out enough to see the box.
I believe it only trains its engine of points based on your 'liking' on articles. To have the system work you have to 'like' content for the recommendations to grow and train.
I've been saying for the longest time that when someone comes up with a social media platform that allows me to actually specify what rule the algorithm follows, I'm in! This gave a small glimpse...
I've been saying for the longest time that when someone comes up with a social media platform that allows me to actually specify what rule the algorithm follows, I'm in! This gave a small glimpse into what that might be like, albeit there are other factors at play such as the pool of content available to browse. Anyway, I enjoy this so much, even in this overly simple form, that I carved out some precious real estate from my bookmarks bar for it.
(I wish they'd call it anything other than a name that made me think it must be another phenomenal Musk invention.)
I actually haven't tried Bluesky yet, but I'm under the impression that the algorithm logic isn't very transparent there either. You just have more options to curate your feed(s)? Which is already...
I actually haven't tried Bluesky yet, but I'm under the impression that the algorithm logic isn't very transparent there either. You just have more options to curate your feed(s)? Which is already great, but perhaps not enough for me to not get siloed into some version of reality that doesn't feel like my own.
My interest is usually not directed at a given topic or viewpoint, but more on the approach/treatment that topic receives, and it's probably not possible to hardcode that into a feed algorithm in a way that would actually serve my needs. I would prefer to see in real time what each like, dislike, express of interest in a topic (and whatever other actions you can perform) does to my results and I'd like to be able to go into the resulting "weights" (sorry, layperson language) and adjust them the way I want, at any time. And ideally, for these adjustments to potentially influence how the algorithm works in the future.
However, this may be an unrealistic dream just because any platform is limited to the content it hosts. My guess is that most platforms simply don't have enough content that I would find interesting for this to work the way I want. Similarly to how dating apps can only show me people who have created a profile - rather than an endless feed full of my dream people who don't actually exist on the app.
Your Following feed on Bluesky is in reverse chronological order. Your Discover feed is more traditionally algorithmic (I mean everything is technically an algorithm.) Your other custom feeds...
Your Following feed on Bluesky is in reverse chronological order. Your Discover feed is more traditionally algorithmic (I mean everything is technically an algorithm.) Your other custom feeds generally the feed maker sets the algorithm.
You can create feeds and set the weighting. Idk about changing it in real time though I don't have the knowledge on the developer side they talk about to know
Thanks for the info! I'm not sure why I don't feel inspired to make an account even though it all sounds good. One factor is probably that I've never been a Twitter user either, not even in the...
Thanks for the info! I'm not sure why I don't feel inspired to make an account even though it all sounds good.
One factor is probably that I've never been a Twitter user either, not even in the very beginning, because it already seemed too.. combative and small minded, or something? Maybe I was just following the wrong people? Maybe I'm also more broadly disillusioned about the power and purpose of spreading around people's half-assed brainfarts vs. content that has faced and stood the test of time. Maybe if there was a platform that would bring these together?
It's not for everyone. I enjoy Bluesky but I don't talk, it's a place for me to learn from lawyers, authors, and other subject matter experts more than it is for me to participate
It's not for everyone. I enjoy Bluesky but I don't talk, it's a place for me to learn from lawyers, authors, and other subject matter experts more than it is for me to participate
This could be fun, I'll start it, post the articles that you opened because you were interested, these were mine: a long list of articles. I opened in the first ten minutes...
This could be fun, I'll start it, post the articles that you opened because you were interested, these were mine:
a long list of articles. I opened in the first ten minutes
I'd enjoy this more if it used standard, not Simple, Wikipedia. An article on turbojet engines was dumbed down too much, the content presented for "science and technology" in an ancient region in...
I'd enjoy this more if it used standard, not Simple, Wikipedia. An article on turbojet engines was dumbed down too much, the content presented for "science and technology" in an ancient region in India didn't include the original article's "Science and Technology" section... I'll stick with Wikipedia:Random.
TIL: Hockey is not the only internationally recognized ice sport.
This led me to discover this hilarious image: https://simple.wikipedia.org/wiki/Sexting#/media/File:Sexting.jpg
At the time of posting, the website seems to be stuck on loading.
Might be your browser, they are using ~60mb of a binary (likely WASM, ie, Web Assembly) to run the model locally, some browsers or network connections have some iffy times loading WASM blobs.
Probably, it eventually loaded on Chrome on Pixel 10, a little faster on Firefox, and faster still on Safari on an iPad.
Ah, yeah WASM support on mobile devices is kneecapped due to the various levels of devices, limited memory support, and that WASM support itself still isn't totally stable (or designed) even on desktop.
This is fascinating to use and see in action. I am fascinated by the negative points for every "heart" I give on some articles - some of these topics are incredible to see go up and down based on my interest. For example: Why did random articles I enjoy end up giving 3022 a science fiction film, so many negative points? In fact, just the reasoning that this system was devaluing it has made me interested - I have never seen it, likely isn't even good, but it is exactly the premise of scifi movies I am interested in - and similar to things I am interested in!
Wikipedia should really take up something like this on the official site, I am discovering so many articles that will destroy the rest of my day. I had a lot planned, now I am planning on even less because I am going down a Wikipedia hole.
If Wikipedia’s goal was driving engagement that would be a great way to do it… but it’s not, and I’d argue it shouldn’t be.
I'd argue that one of the things that's kept Wikipedia out of the Enshittification whirlpool is the lack of interest in engagement.
Hank Green recently posted a video going into this very topic: https://youtu.be/9zi0ogvPfCA
Wikipedia's goal is really to catalogue and explain the universe, from a human lens, for the betterment and sake of knowledge and human existence. I would say that an interface that draws more people in, to encourage research, even through a simple addon like this (they kind of do already, but through daily editorials), and they could really tap into some interesting knowledge on human interests, and inspire, create, and cultivate a wider culture to become editors (we need more of them!) and better thinkers.
We generally need more researchers in the world, and more platforms that feed the need for social feeds, with something beneficial for the mind. A bicycle of the connected mind, if you will.
Perhaps Wikipedia is the quality it currently is because the wrong kind of editor self select out of it.
Maybe more engagement would dillute the quality contributions among a sea of low effort ones.
We don't really have many good things like wikipedia, we should be carefull with it.
Don't want to further inflate the ego of editors with impression metrics...
How do you see points? Does this only value things that you open or like? Why would you like something?
I see how I opened one article and it started giving me a bunch of related articles I have no interest in; how do I know if it's figuring out anything else about me?
If you are on mobile, it looks like they hide the stats box. Mine off to the right side says:
should be noted, I did not 'like' or decide not to like anything about "Category:Isle of Wight".
If you want to see your stats on mobile, set your webpage view to 'desktop mode' and it will zoom out enough to see the box.
I believe it only trains its engine of points based on your 'liking' on articles. To have the system work you have to 'like' content for the recommendations to grow and train.
I've been saying for the longest time that when someone comes up with a social media platform that allows me to actually specify what rule the algorithm follows, I'm in! This gave a small glimpse into what that might be like, albeit there are other factors at play such as the pool of content available to browse. Anyway, I enjoy this so much, even in this overly simple form, that I carved out some precious real estate from my bookmarks bar for it.
(I wish they'd call it anything other than a name that made me think it must be another phenomenal Musk invention.)
Thanks for posting!
Does Bluesky fit the bill at all? I don't feel like I've had anything pushed on
myme that I didn't request. Admittedly I don't use it overmuch.I actually haven't tried Bluesky yet, but I'm under the impression that the algorithm logic isn't very transparent there either. You just have more options to curate your feed(s)? Which is already great, but perhaps not enough for me to not get siloed into some version of reality that doesn't feel like my own.
My interest is usually not directed at a given topic or viewpoint, but more on the approach/treatment that topic receives, and it's probably not possible to hardcode that into a feed algorithm in a way that would actually serve my needs. I would prefer to see in real time what each like, dislike, express of interest in a topic (and whatever other actions you can perform) does to my results and I'd like to be able to go into the resulting "weights" (sorry, layperson language) and adjust them the way I want, at any time. And ideally, for these adjustments to potentially influence how the algorithm works in the future.
However, this may be an unrealistic dream just because any platform is limited to the content it hosts. My guess is that most platforms simply don't have enough content that I would find interesting for this to work the way I want. Similarly to how dating apps can only show me people who have created a profile - rather than an endless feed full of my dream people who don't actually exist on the app.
Your Following feed on Bluesky is in reverse chronological order. Your Discover feed is more traditionally algorithmic (I mean everything is technically an algorithm.) Your other custom feeds generally the feed maker sets the algorithm.
There's info here
Algorithmic Choice with Custom Feeds - Bluesky
You can create feeds and set the weighting. Idk about changing it in real time though I don't have the knowledge on the developer side they talk about to know
Thanks for the info! I'm not sure why I don't feel inspired to make an account even though it all sounds good.
One factor is probably that I've never been a Twitter user either, not even in the very beginning, because it already seemed too.. combative and small minded, or something? Maybe I was just following the wrong people? Maybe I'm also more broadly disillusioned about the power and purpose of spreading around people's half-assed brainfarts vs. content that has faced and stood the test of time. Maybe if there was a platform that would bring these together?
It's not for everyone. I enjoy Bluesky but I don't talk, it's a place for me to learn from lawyers, authors, and other subject matter experts more than it is for me to participate
Why is this flagged as 'nsfw'? Unless it's just dangerously distracting?
The site itself says it may well surface nsfw material.
I got tentacle rape on the first page :-/
Including image
As it says on the page, it will show Wikipedia pages with nsfw content without any warning.
Wikipedia doesn't have a filter for explicit content, and I didn't want to run into a Wikitok issue again.
I got a page on circumcision (complete with image examples) very early on, so it seems like an important tag.
Really cut to the chase with that one!
This could be fun, I'll start it, post the articles that you opened because you were interested, these were mine:
a long list of articles. I opened in the first ten minutes
https://simple.wikipedia.org/wiki/Pneumatics
https://simple.wikipedia.org/wiki/Photomask
https://simple.wikipedia.org/wiki/Stabiae
https://simple.wikipedia.org/wiki/Combo_XLR/P10
https://simple.wikipedia.org/wiki/Technosexuality
https://simple.wikipedia.org/wiki/A%C3%AFr_Mountains
https://simple.wikipedia.org/wiki/Cybernetic_immortality
https://simple.wikipedia.org/wiki/Nara_period
https://simple.wikipedia.org/wiki/History_of_Manipur
https://simple.wikipedia.org/wiki/Technological_convergence
https://simple.wikipedia.org/wiki/Crystallography
https://simple.wikipedia.org/wiki/Uniting_for_Consensus
https://simple.wikipedia.org/wiki/Gecko_(rendering_engine)
https://simple.wikipedia.org/wiki/Emperor_Go-Toba
https://simple.wikipedia.org/wiki/Solar_wind
https://simple.wikipedia.org/wiki/Sarco_pod
https://simple.wikipedia.org/wiki/Undersea_cable
https://simple.wikipedia.org/wiki/Stupendemys
https://simple.wikipedia.org/wiki/Turning
https://simple.wikipedia.org/wiki/Science_park
https://simple.wikipedia.org/wiki/Innovation
https://simple.wikipedia.org/wiki/MOS_Technology_6510
https://simple.wikipedia.org/wiki/Russian_Soviet_Federative_Socialist_Republic
https://simple.wikipedia.org/wiki/Lattice
https://simple.wikipedia.org/wiki/Sounding_line
https://simple.wikipedia.org/wiki/Phenomenon
https://simple.wikipedia.org/wiki/Balloon_(aircraft)
https://simple.wikipedia.org/wiki/Household_hardware
I'd enjoy this more if it used standard, not Simple, Wikipedia. An article on turbojet engines was dumbed down too much, the content presented for "science and technology" in an ancient region in India didn't include the original article's "Science and Technology" section... I'll stick with Wikipedia:Random.
TIL: Hockey is not the only internationally recognized ice sport.