I versus Algorithms 2 – Who moved my home page?

I woke up interrupted on a Saturday morning with my wife furious at our sons.

“You are not getting my phone, again.”

Both the sons claimed innocence while shifting blame at each other; in the crossfire my desire to get additional sleep got a quiet burial.

My wife was looking for some breakfast ideas on Youtube and found her home page filled with Cartoon videos playlist, as long as she could scroll down. 

“This isn’t how it was a month back; I had my cookery, comedy and news playlist. I gave the phone to them for cartoon drawing tutorials, and here is what I get.”

Greater surprise awaited my wife, when she moved to check her Google Updates list and that also had majority items on new releases from the cartoon studios to fun facts, fan sites et al.

“Blame it on algorithms, not them” I tried to reason with a smile.

The cat and mouse game of algorithms

Let’s look at the world of algorithms with a rather simple abstraction – the cat and mouse.

The mouse doesn’t live to be cat’s food; neither is eating mouse, the only purpose of cat’s life.

However, their lives are interconnected with survival as intersection point.

The cat tries to get smarter at catching, and so does the mouse in eluding in their own cognitive ways.

To go few levels deeper into their world, the cat doesn’t randomly hunt its prey; it maps the possible sites, paths to the sites including where the chances of finding mice is higher, and if it can throw in some inducements.

The mouse goes about its own life finding food, it’s shelter and a consciousness against potential traps, or chance encounter with cats.

Similarly, our increasingly digital lives have footprints; no matter how random or, unique they are connected with objectives – to be informed, entertained, undertake commerce, interact or, ideate.

These footprints act as breadcrumbs that drive algorithms developed by companies, platform players with almost the same objectives. These are the intersection points.

As the footprints increase, algorithms get trained better, get smarter.

The recommender algorithms in Youtube with the initial data of cartoon making tutorials, figured out that the user loves cartoons and loaded some initial lists of favorite characters. My sons started with the many facets of cartoon making and later changed track to watching the many unseen videos of their favorite characters.

The intersection point changed to a full area of overlap, until that Saturday morning!

The evolution

It has been almost 25 years since we dived into a more networked world.

From music playlists, news, commerce to even social networks there is a list that’s unique to us. The order of news that I get and you get is different, even to the extent of facts and opinions around the same news.

This customization of epic scale is not just design of algorithms but carry objectives of their developers, organizations with goals from economic, scientific, social to the political.

As network and user base grows in scale and complexity, algorithms increasingly supplement human decision-making, to a point of replacing human decision making.

It is as good as how automobiles and aeroplanes supplemented and augmented human travel, or electron microscopes allowed us to penetrate beyond capability of human eye to improve understanding at molecular level to the tools that we require for accelerated vaccine research.

The growth of algorithms is perhaps not just an outcome of computing power and massive data, but what Galileo in 16^th century had insisted that mathematics is the language in which God has written the universe.

Scientists over the centuries from Aryabhata, Newton, Copernicus to Einstein made pathbreaking advancement in developing human knowledge without computing resources with mathematics, and influences from science and philosophy.

Hence, algorithms are the evolution of centuries of human quest to understand the universe by building mathematical structures to our physical world and behaviours.

The data deluge and algo troubles

An estimated 294 billion mails, 65 billion Whatsapp messages, 5 billion searches 500 million tweets and petabytes of data are created across Facebook to major platforms and social networks, everyday.

4.6 billion people accounting for 60% of world’s population use the internet, with objectives.

As governments step up digital connectivity and surveillance, companies digitize more and 3.8 billion users on social media intersect, there is friction at play.

The friction is not in the algorithms that decode, predict or recommend, but what they are deployed for.

To increase its sales if a company follows your digital footprint through apps and websites to plant its irritable ads without permission, then it has no respect for your privacy.

As some governments have developed surveillance to restrict specific freedom or drive a narrative, have also led to creation of moderating algos that muffle free speech.

Now, let’s look at our own “Top picks for you”, “From your past orders”, “For you” and “Today’s stories” lists across apps and sites. What do they tell about our choices, about us?

Shallow, impulsive, hurried, deal-hungry, attention-seeking or, nuanced, rigorous, curious, exploratory and creative?

Time to reset purpose

When we look at the jaw dropping developments in basic sciences over the course of human history, a lot was achieved with spirit of enquiry and with elementary tools, from the standpoint of what we have today.

There is a serious need to re-ignite, recognize and reward such quest, such progress.

There is a case to use more filters, deny app permission and discard random feeds so that we create space on internet that helps us with our objectives.

As we work to grow our intelligence as a society with integrity, respect for rights, champion new knowledge and make this accessible, these virtues will flow into our algorithms.

How do your recommendation lists look?

_________________________________________________________________________________________________

May also read - I versus Algorithms 
http://bluepeepal.blogspot.com/2019/02/i-versus-algorithms.html

References:

https://www.britannica.com/biography/Galileo-Galilei

https://www.weforum.org/agenda/2019/04/how-much-data-is-generated-each-day-cf4bddf29f/