The Filter Bubble
The hidden rise of personalization on the Internet is controlling--and limiting--the information we consume. In 2009, Google began customizing its search results. Instead of giving you the most broadly popular result, Google now tries to predict what you are most likely to click on. According to MoveOn.org board president Eli Pariser, this change is symptomatic of the most significant shift to take place on the Web in recent years--the rise of personalization. Though the phenomenon has gone largely undetected until now, personalized filters are sweeping the Web, creating individual universes of information for each of us. Data companies track your personal information to sell to advertisers, from your political leanings to the hiking boots you just browsed on Zappos. In a personalized world, we will increasingly be typed and fed only news that is pleasant, familiar, and confirms our beliefs--and because these filters are invisible, we won't know what is being hidden from us. Our past interests will determine what we are exposed to in the future, leaving less room for the unexpected encounters that spark creativity, innovation, and the democratic exchange of ideas.--From publisher description.
Stats for nerds
3recommend
13mentions
13people
2011–2025years
23% recommend69% neutral8% critical
Mentions per year
What Hacker News says
- > These services don’t control ‘the’ flow of information, they are entirely additive to what we had before.
They very much do control the flow by "personalizing" the content people see and don't see based on how likely somebody is to click something for ad-revenue. That's the direct result of the "capitalist nature" of these platforms; If they can't monetize it, it's of no interest to them, so they try to capitalize on everything, even on users biases.
Surfacing news about how refugees are not raping everything in sight won't do much for a user who has a very clear anti-refugee bias, such a user won't be very likely to click such news as such they won't be surfaced to the user. But if you surface news about how Germany is supposedly on the brink of collapse because refugees are roaming free as raping herds, to that very same user, then you will most likely just have generated ad-revenue.
This doesn't just extend to what kind of news get surfaced, but through the whole system: Suggestions for friends, groups, and whatnot all take into account the individual users known "leanings", in essence creating echo chambers where only people of similar views interact with each other. While at the same time blending out any alternative/controversial view (for that group) by not even surfacing them to in the very first place, as they won't be very likely to click on them and as such less chance for ad-revenue.
That's why so many people are so utterly convinced of their positions: Their social media feeds give them constant reinforcement for their position by only surfacing news, people, and organization who share these position.
Eli Pariser's book "The Filter Bubble" has some very good (and scary) insights into this dynamic.
- Great TED talk on the "Filter Bubble": http://www.ted.com/talks/eli_pariser_beware_online_filter_bu...
- Isn't this a quote from "The Filter Bubble"? [0] Fantastic read by the way, for anyone who is interested in (or terrified of) behavioral advertising.
[0]: http://www.amazon.com/Filter-Bubble-Personalized-Changing-Th...
- Hello, untruths!
>The term filter bubble was coined by internet activist Eli Pariser circa 2010. In Pariser's influential book under the same name, The Filter Bubble (2011), it was predicted that individualized personalization by algorithmic filtering would lead to intellectual isolation and social fragmentation.[1]
Ctrl+F for "carr" on the "filter bubble" wiki reveals two instances of "carrots". Similarly, Ctrl+F for "bubble" on Carr's Wiki returns 0 results. The book Carr is most famous for was based on an article that he wrote two years prior to Pariser coining the term, and was published the same year the term was coined. His following two books also don't seem to be riding on some kind of "filter bubble" coattail (I mean, unless you've found a way to connect his next book, a 2014 release about automation, to filter bubbles!). It was Pariser who published "The Filter Bubble" a year later.
Essentially, what you've done is written this piece off entirely based on a false assumption, and suggested that others do the same.
- New Public (https://newpublic.org) | REMOTE in US | Full-Time or Part-Time
We are hiring a software engineer to work on Public Service Incubator.
New Public is working with a group of Public Service Media organizations across Europe and North America to create a new publicly-owned public-service online platform that is designed to naturally encourage productive discourse between people who see the world differently.
Normally creating a new online social space is super hard (network effects and all that) but our partners have the credibility needed to make a new space high status, and the power to direct the attention of millions of people. So we think we actually have a decent shot.
We've built a prototype that our partners and test users are pretty excited about, and now it's time to turn it into a real high quality product.
We are particularly interested in people with experience shipping social products, an ability to work anywhere in the stack, and a deep interest in finding ways to help people who see the world differently understand each other and work productively together.
Stack: React (Native) | Node.JS | NoSQL | OpenAI
New Public is a non-profit that was founded by Eli Pariser (The Filter Bubble, Upworthy, MoveOn) and Talia Stroud (Prof at UT Austin) with the vision that we should focus not just on how to reduce the harms of online spaces, but how to design them so that they make good things happen. Public Service Incubator is one of our flagship projects.
- >Is there an established "kittens to terrorist" pipeline?
Yes, its what Eli Pariser described as a Filter Bubble, first identified in Google Search results. I wonder if there has been some code sharing going on between Google and Youtube, or if team members have jumped ship?
https://www.ted.com/talks/eli_pariser_beware_online_filter_b...
https://www.theguardian.com/media/2017/jan/08/eli-pariser-ac...
https://en.wikipedia.org/wiki/Filter_bubble
Some youtube holes can be good but considering the untapped potential of Youtube for DJ mixes as background music for coders, I find the youtube hole incredibly frustrating.
I mean I've only just found this DJ and its pretty decent mix and her previous work is quite good as well, but considering she is Ukrainian, I wonder if this is Youtube's pro Ukrainian political policy in effect? https://www.youtube.com/watch?v=oYC8KkNIkZk
I found this decent mix (https://www.youtube.com/watch?v=5vHRUsP20dQ) only after listening to a couple of 145bpm German Underground techno mixes.
I mean what were the algorithms thinking?
The only link I can see is its from the Boiler Room in Berlin and Berlin is where you can find some decent German underground techno, plus the time of day is a factor, ie trying to slow down someone's Dopamine stimulation at 4am in the morning?
But if you listen to some of the more established DJ's like Boris Brecha, Carl Cox, David Guetta, Adam Beyer or brands like Creamfields, Global Underground, Tomorrowland, Resistance, Transmission, Hooj Tunes, Essential Mix, you'll get stuck in youtube holes left right and centre and the quality tends to get worse the longer you listen. I mean their algorithms dont know how to curate DJ mixes, just like most people cant take people on a journey of aural delights when DJ'ing.
One other thing I have noticed with Youtube is bank holidays in the UK, they tend to promote borderline illegal car activities, things like cruises where people lose it leaving the cruise or speeding videos, more so on August bank holiday and the last May bank holiday. I'm left with the impression Youtube is trying to incite illegal car activities, so maybe the NY Times and the Guardian do have a point!?!
Related
Probability Theory
77 recommend · 193
In the plex
35 recommend · 100
A new kind of science
44 recommend · 388
Filters against folly
3 recommend · 4
Elixir in action
48 recommend · 72
Introduction to the theory of computation
66 recommend · 96
Decoding the Universe
4 recommend · 4
