The unplanned nature of this post is part of that effort.


It’s been said that no one knows how the YouTube recommendation algorithm works [*TS], or maybe what exact number determines if you see these words on a search engine.

What it promotes, what it favors, what it despises. No one knows.

And yet, people rely on these recommenders more and more around the world, like me, clicking for a brief period of entertainment during my commute, or like thousands of social media influencers, trying to make a living out of it[*IA], we welcome it into our lives at varying paces.

It’s been baffling me, with how isolating and self-amplifying the Internet sometimes has already become, long before the age of automated content delivery, how humanity will fare, into such an embrace by all those automatic recommendations, which may or may not work in our favor.

Content-delivering AI’s. I understand the need for them, at least on paper. We generate data at a rate never before seen in the history of our civilization, and for anything to get anywhere, we will eventually need to employ the help of our machine friends. I do not question the usefulness of automatic filters and such, but struggle to put my trust in them as they currently behave.

As a consumer, what you liked will decide what gets delivered to you next; and what you dislike naturally falls outside your sphere of attention, which is limited to begin with.

Good idea!

Good idea?

It all began sounding simple and fair.

A computerized curator, another great idea with potential to be corrupted by execution to be sure.

As a recent example, the one that prompted this essay, a video that I watched that reviewed a recent film negatively, soon meant my YouTube feed got flooded with critical Video-logs (vlogs) of said film. Here’s your watch time quota, I almost heard the app say to me, with no regards for factual consistency, or if it is my true intention to take the views exclusively from one side. For netizens that are satisfied with or trust front page news, the majority perhaps, the current dynamics is more than the cause of an illusion of consensus, and silencing of opposing views.

Connected, we divide, spontaneously.


It’s been quite a while that I keep telling myself not to press that ‘follow’ button for information that I actually care about the most. I started stockpiling PDFs of important webpages instead of Facebook ‘Save for Later’s, I made it a rule not to stay on ‘Recommended for You’ or ‘Trending’ for too long every day, and I find switching to DuckDuckGo as my search engine a step towards the right direction.

To be fair, these changes happened partially because the people I do wish to ‘follow’ aren’t even on social media any more. That aside, I do feel an increasing need to ‘seek my information offline’, with the associated difficulty of doing so online growing in tandem.

All these sound like backwards progress, don’t they?

When my relatives consistently send me news articles from one website, or when a person with whom I debate online quote all their supports from one author… I sweat a little.

There have also been office talks about my fellow classmates ‘beating the AI’ by searching for intentionally weird or different things… They have a point.

As things currently happen, social, entertainment or news content that automatically aligned with my prior conception, has caused me great discomfort.

When you stare into a homogenized remix of your pre-formed views and knowledge, only those will echo back, too, along with a false sense of impartiality, and the danger of losing track of what you do not know.

I prided myself in creating content more than consuming, I still do, I hope, but worry how much I too contribute to the Automatic Recommendations for humanity.

There’s no hint that help will come from elsewhere to save us from ourselves. [*CS]

And I suspect staring into our own reflection on a still pond is the way either.


I often feel a related difficulty while I write, that I should elicit a certain extend of doubt in my readers, about them, and, certainly, about me. Keeps you on your toes.

While I do hope this article gets picked up by an auto-recommender so it could reach you, I do hope this dynamic dies away soon, and you are pointed here via random wandering or word-of-mouth.

Here are some other ways that I use, apart from those vague points in the chapter above, to live outside the influence of auto-recommendation algorithms.

. Random page .

Every Wiki should have this standard equipment. Here is the one for English Wikipedia:

I have been reading random articles during lunch since 2017, and compared to the talk-material it endowed me with, what I appreciate the most is the constant reminder of the extent to which I can overestimate my knowledge.

A related wiki-activity is the famous Wiki Race, in which people compete to navigate between two random wiki articles by clicking links only.

. Surprise Me! .

A similar button in WolframAlpha, gives you random quantized facts.

. Having Cool Friends .

I’ve been fortunate to never lack this valuable treasure of life since middle school. Online and in writing, we take different views and differences hardly pushed us apart.

Another related side is quitting Facebook … or actively post original things to it (not so original that they steal your intelligence property rights).

I will remark about Facebook’s news feed at another time. I use the plug-in called Feedless to disable it on my iPhone, so I don’t really see it that often now.

. D.O.N.G. .

“Do it Online Now Guys”, Micheal Stevens’ another YouTube Series.

These videos are one of the quick means into the fun, informative, and epistemically challenging side of the Internet.


Join my journey …

I will keep a critical outlook, and never satisfy with a false sense of accomplishment. The previous sentence is just a glorified way of saying self-depreciation, but it works in frontier science :), I think.


[*IA] 4 Ways to Beat the Instagram Algorithm,

[*TS] Why The YouTube Algorithm Will Always be A Mystery, Tom Scott,

[*CS] Quote from Carl Sagan’s Pale Blue Dot Forewords.

[*QQ] II is based on ideas summarized from multiple discussions between me and my friends, as well as a Chinese social media post.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.