Funnel Vision
Oh no! Another article on AI and machine learning…
Well, hello people and welcome to this article. You might wonder if I am a highly intelligent AI trying to convince you that I am not, and with all fairness, at this stage, you can’t be sure. Do we live in a simulated reality (article)? Will AI ever overtake humanity (article)? Are we close to technological singularity (article)?
We won’t touch on any of these subjects but rather explore the effects of AI on the way we consume digital products and how personalised search and results can drastically affect the way we digest media online (and sometimes — more and more often — offline).
Make it personal
Personalisation has been around for a while and became extremely prominent in the era of smart devices. As the number of smart devices increased, so did the apps and content that these devices were filled with. Although we, the end users, benefitted from this over-saturated, globalised market by having loads more to see and read, in the same time our senses were overworked, our brains became tired and our stress level is at it’s all-time highest. Our fear of missing out combined with all the user-friendly content provided for us has created a combustive cocktail.
I won’t even touch on how feeding into our greedy needs as an over-consumer is just a terrible idea together.
The next big step in the evolution of content providing should be to start educating people on how to consume information online.
To help them transition from passive consumers to self-conscious online personas. We are not on this path and there isn’t much effort made in this field either.
Recommend and Conquer
Okay, so personalisation is big and it is very important to elevate the user experience to an unknown height. There are commercially available AIs out there that can personalise a website in real time (video). But there are also extremely popular applications out there with similar technology used by millions as we speak.
I guess you have heard about Netflix, right? The pinnacle of online video streaming? If you are a person interested in such things, you might have also read articles or seen videos on how Netflix personalises content to every single user differently (video). And in itself, it’s not a bad thing. On the contrary. Having an experience tailor-made to our unique needs can only result in great customer experiences.
What about Spotify? Online music streaming? Do you use it? I do. Often. One of my favourite things in it is Discover Weekly, a playlist with new artists and songs I never heard about but like very much (apparently). It is also fascinating to know how it’s being brought to life(article). And there is also Release Radar (article). Similar idea and equally as addictive.
We have great content providers and great content curated just for us. What is the issue?
Enter the Magic Funnel
The issue is what this content curation makes us do or more like what it makes us see.
By looking at videos and listening to music based on our recommendations we are positively reinforcing algorithms in their work, we unintentionally create our own filter bubble (ebook). The more you rely on recommendations the more it will narrow down the things you can encounter online to a certain niche that is — or more like was — your taste when the algorithm started tracking it.
Let’s say you were into fidget spinners a year ago and were watching plenty of videos on them. Now, a year later, you are not that keen on them anymore but since you have been watching videos related to the topic quite heavily in the past, it still crops up in your homepage on YouTube. Since we are getting lazier by the day, instead of setting out and searching for new, interesting topics on any video streaming platforms, we just flick through the first few hits and chose one of the first ten videos.
This leads to content funnelling: the more content we consume based on our recommendations, the more limited the variety of content we are seeing.
AIs (algorithms) —in pursuit of maximising our engagement with their product — tailor our view of their immense catalogue of content down to a niche selection based on our demographical, political and geographical parameters. Do you want to see what Saudi Arabian science bloggers are up to nowadays? Too bad, but hey, here is yet another fidget spinner video.
Censorship of the Self
As I mentioned before, it takes a lot of effort to escape the vicious cycle of recommended content. The more overwhelmed we are with information, the more we tend to lean towards recommendations. It also creates another issue which is equally as dangerous and certainly limits our view of information online.
If on an Android device you open Google Chrome, it will — after a while — show you recommended articles (in a form of an endless scroller). You happen to like this feature, it selects relevant articles you have searched online before. But wait, you — or someone using your phone — have opened an article about Love Island, a popular reality TV show. Are you not interested? Too bad, articles related to Love Island will keep on cropping up in your feed.
You decide to outsmart the algorithm and make a decision to only click on articles that are related to subjects you really like. There might be other, interesting looking articles on the list, but you won’t read them because by choosing said articles you might taint your future recommendations.
This is self-censorship: you stop consuming information that might have a negative effect on your future recommendations.
Sad as it is to admit, but by combining the effects of these two phenomena, we end up seeing a very narrow segment of available content on the internet. Content funnelling may be convenient and self-censorship might be effective to some degree, but at the end of the day we lose out on a lot of available information.
Remedies and Suggestions
- Consume information on a variety of platforms
- Combine online and offline information
- Cross-reference your information in different media
- Keep your mind open!
