Power resides in the eyes of the beholder
The curiosity, the cat and the killer (Ep 1)
Welcome to episode 1 of my new short story “The curiosity, the cat and the killer”. Power is a tricky possession. Dynamics of which, have been prevalent in human society since their genesis, in varying forms and scales. People have resorted to extreme ends to come to power, to continue to be in power, or sometimes just “flaunt” power. And when this addiction is coupled with a level of sophistication, collectivism, and planning, governments are born.
2030 — a decade-plus, and China has become the global source of inspiration for what it means to “own” a government. The western countries have given in and started employing ideas from Chinese political strategies into their respective governments. “Manipulating people is a right reserved by a government” is an unsaid aphorism the Chinese government strives for, consequences of which are deep-rooted across domains. And technology has become the primary medium to burgeon this cause. The persistence of such anarchy for a long duration has resulted in organizations like Alibaba (Amazon), WeChat (Facebook + WhatsApp + Venmo), Baidu (Google Search) and Yoaku Tudao (YouTube) now being owned by the government. A conspiracy theory that these companies were “friends” with the government all along and just put on shows after shows before us to trap us completely into a network of manipulable brains — is a prevailing dark folklore. Those in power saw this as a masterful way of creating a self-funding, crowd-sourced, gaming-industry of playing mind-games. The ability for government operatives in authoritarian countries to systematically poison the information environment in democracies with minimal effort was no big deal.
Using data the government now owns, it can fabricate a digital representation of the entire physical world in terms of experiencing life. Like, the calendar tells your routine, your WeChat payments tell about events in your life, maps tell us where you are and social media tells us who are you interacting with, what you search tells us what you are curious about off-late — every life information being translate into a time-series of JSONs on the computer, each of which extrapolates to the mental and physical state of the person(s) it represents. Building on these representations further, analytics projects have been going on for years and the time series could now be “predicted,” by an algorithm, making your life “predictable” quite literally given a little of history. With almost all the people in the country now connected with these companies, it has vested supreme power in the government. The government used social media to plant seeds and emanate its authoritative ideology among people. Spies, both robots and humans were hired to watch over the internet traffic and report/brainwash those raising a voice against it. They wanted to make their credo so deeply rooted that it felt organic and conscientious to the people. This by design would eliminate protesters as they would be disdained by the rest. Snooping through automation devices and mobile phones is now old school and commonplace. Paradoxically, industry engineers are now working hard to access private data by leveraging the faculties of The Great Fire Wall (GFW), while hackers are being hired by activists to breach networks and reach delicate internal resources of the government for a coupe. The intellectual battle between the government and Virtual Private Network (VPN) providers is still on, attempting to mask-unmask outgoing traffic. Giant conglomerates like WeChat, Yoaku Tudao are broken into several small teams working independently, their collective data aggregated and shared, after they were taken over by the government — a flatter hierarchy to warrant more control. Managers and product architects are recruited by the Chinese government to influence companies to ideate, design and implement products which would instill relentless power in controlling people as individuals, as a family and as a group with technology being their indispensable tool. Sometimes engineers would not even know where their day to day work is going to be used.
Ying Yu Ho, a 29-year-old workaholic geek, was a ninja-like software engineer in one of the leading teams at WeChat in Shenzhen, the Silicon Valley of China. Yu’s team was devising predictive models on the messages being exchanged in large groups, one of the dirty teams in the Chinese tech world. The government had decided to stop encrypting messages within groups of twenty or more people and Yu’s job is to analyze and predict the semantic sense of the mood in a group and file an alert with the government when necessary. Ying Yu was among the fastest in execution. He liked his job but his unrelenting curiosity in the product-space and incessant urge to try something new often had kept him so distracted that he could never fall in love with his job, or with any job in general. “What problems are people facing? What new sources of data and services are available? Can some of this data be modified or even captured to solve a problem? What aspects of communal interaction and events can be learned from public data at hand?”, product-level questions like these were never out of his head. He would keep building newer prototypes (or many times just low fidelity hypothesis), but often not take them to fruition, because sometimes they did not make a consequential difference or many times he would just get bored. His curiosity would seldom turn into live actions and only a fraction of those actions would end up with any result, be it success or failure.
“What people are talking about?” , “How are common people reacting to a decision?”, “Is there a trend which suggests the emergence of a new activist group?” were common questions his team dealt with. Hackathons were Yu’s favorite events, where he could superficially prototype a solution and get validation in a few days. He refrained from getting to the depths and details and was happy just hacking solutions across the breadth. As a result, Ying Yu had gotten very good at developing a broad range of high fidelity prototypes quickly, which was a great asset to his team. Taking part in hackathons internal to WeChat was something he looked forward to every year, and this time was no different. With the access and exposure to voluptuous (read suspicious) data, exhaustive tools, cutting edge technologies and constant support he would get, there were limitless things he could try. And given the standards the company had achieved over years, more often than not, any new technology would be easy to pick up and would work right out of the box without much tweaking, even when integrated with an external system. The hackathon coming up was called Data Confluence, where one could potentially use any data (a small snapshot) which now included everything from social network content to census data, from videos viewed to search history, from shopping statistics to blogs read and build a product around it. With two days left for idea submission, Yu was pondering sitting at his favorite thinking place — the toilet seat on his floor at work. He was skimming through the learn-on-the-potty article for the week, stuck on the back of the door and discovered that WeChat had been working on a service which provided a mathematical representation of a person, based on his profile, likes and dislikes, followers and followees, browsing history. Over the past decade, although the tech in China was not successful in creating physical intelligent robots (like a prototype Jarvis), companies and research institutes in different parts of the country had established immense success in the mathematical modeling (an n-dimensional representation) of a person at-least i.e. even though an Ironman’s autonomous super-powered assistant Jarvis remained still a fantasy, a fraction of Doctor Strange’s brain ability to predict and evaluate future possibilities to choose the best outcome had become a reality. Ying Yu was very intrigued by the idea of converting a physical person and his life in a continuum of Cartesian space to a vector in some information space.
“People are so fake!”, yelled Chunhua (Ying Yu’s friend) scrolling on the chat of a dating app while they hung out for dinner with the rest of their group. “Tens of apps, hundreds of features, thousands of guys swiping right on autopilot and a bunch of girls mindlessly chasing the elusive virtual creamy layer. Nobody has still managed to crack the supply-demand problem here. Guys suffer from the dearth of approvals and women are overwhelmed in filtering out genuine ones from a pool of basically all the guys, because come on, who swipes left”, she babbled looking at Ying Yu, who was too busy playing on his phone. “You gotta take some chill”, he replied with least interest although he heard every word of what Chunhua ranted. On his ride back home, Ying Yu brooded over Chunhua’s problem of people not being genuine. “People fake mostly to create a wall of glamour, only to give others FOMO, or out of peer pressure. Happiness, self-security and emotional sufficiency are seldom found within these walls” he thought to himself. “What is the one place where people would not lie? What is the one place people have no filter about their identity and they would not have to wear a mask? Where is the one place where people cannot have any insecurity and have to be themselves while interacting? What interaction is more a routine and not occasion driven?” he pondered as he gawked on the street lamps passing by.
“A search engine! A web browser!” eureka-ed Ying Yu out loud in the cab. “People do not lie while they look for something on the internet, be it any form of content, people might choose to conceal their identity (anonymity and incognito) but would be themselves while browsing the web, lest they would not find information”, he confirmed to himself in his head.
Ying Yu was interested in giving a shot at this problem by leveraging what he had read previously. He recounted this and brainstormed with a senior member in his team, Youngjin and they both concurred it would be a useful application to solve the fakeness problem in apps that people complain the most about online dating. Extracting attributes like things he/she searches for, music (and videos) he/she listens (and watches), shopping trends, websites visited — they believed a fairly true representation (vector) of the person (his traits and his interests) could be obtained, specifically customized for the use case of finding a date. And if some form of a distance measure could be devised to compute by how much the two vectors vary, they believed it could be a valid illustration of the likelihood that the pair would potentially be a good match. Two days of toiling, involving orders of feature engineering (picking the right contributing attributes for matching) to tweak the existing model and numerous jugaads of biases later, they extracted features for a few thousand people. They modeled the matching problem as an edge weight prediction problem. Representing every user as a node and the edge weight as a representation of the “distance” or “matching coefficient” between the pair, they trained a pipeline over the next 2 days. Even after overfitting, the model was only OK, but good enough to demonstrate the idea. With this not-so-accurate solution, they anyway presented their idea showing how two potential partners would have been recommended by the model and how it was a true positive because the couple got married a few months ago. They did not win a prize perse, but the idea sure garnered many eyes from the executive boards at WeChat and Baidu. Ying Yu was just happy from the learnings he had in the new domain. After a few days, on a lousy afternoon, his yawn was dismissed by a calendar invite from the VP of the organization regarding the hackathon project. In the meeting, Ying Yu and Youngjin found out that the VP was interested in investing more time into the project along with a small team of three helping engineers. The VP believed that the idea encompassed a much bigger potential for people manipulation (the govt’s endmost credo). A collective wide grin — full of depravity and a shake of hands — full of approval concluded the meeting.
