September 12, 2020

Down to Rabbit Hole

Down to Rabbit Hole

Rabbit Hole: place for a wonderful or troubling surreal experience, usually used in "going down the rabbit hole" or "falling down the rabbit hole".

The phrase stems out Caroll's 1865 Alice's Adventures to Wonderland. In today context and with Internet, down to the rabbit hole refers to an extremely engrossing and time-consuming topic. More on

And if you have time for podcasts, listen to a series by The NewYork Times about Rabbit Hole in context of the information was used in the Internet where users in trenched topics, and one step led another, the users starts to believe to an extreme view. Down to Rabbit Hole is dangerous when politics is blended in and resulted in physical harms to individuals, to fellow citizens, by those who participate in legitimate means such as election, campaign, protest but with a bias information that inadvertently leaves a long-term damages to people liberty, the values of union, and democracy.  Trenched topic is the key here, that is a "cluster" of relevant content with the similar tone, genre, topic to the watched video, for example.

Underneath, a term such as "recommendation system", - one or a collection of algorithms built by giant Internet content providers such as Facebook, YouTube, Netflix, Amazon and others to first understand your preference and provide more of the content you likely to consume and spend the time on. A simple thought is ranking content by your reference, so that by presenting those videos or movies, articles in the next windows, a higher chance you continue spending time on the page. You save time searching a similar content, the provider benefits from a loyal consumer. In a business case, this is a resounding win-win case.

Around 2016, Target, a supermarket chain in the U.S., operated an algorithms to target potential customers by collecting contents that customers searched on their website. The algorithms, for example, can predict a particular customer likely to be pregnant and for how long, and subsequently sending mail to promote products after-birth such as diapers, baby powders. In 2020, most of the sale platforms has mastered to track user's preference or interest items. What was the last time you search for a pair of shoes, and when you browser on different pages, great deals on shoes just pop up, patiently waiting your clicks.  Have you watched the news lately about shootings in the U.S., and somehow after one massacre occurred, the police looked into motif of the shooter by looking through the search query one did over Google and such. Everything could be gone except the digital trait you left in some database sitting there until one day someone figure out what to do which those data.

Smart engineers and company use the preference recommendation algorithm not only to sale physical stuffs, but also the idea, perspective, and often these days, misinformation. End goals are varied but underlying engineering is identical. If I know you have searched for a voucher to discount a hotel price, I could guess that you would like to have other travel accommodation such as skin care, taxi, where to eat, things to do in that area. What were the last time, you booked a hotel and then a couple hour later you have a phone call asking if you need a lift, or that agent can send you a list of articles to organize your trip. This is benign. An offer of information for your gain of time. But what in the market has only a few options, or they can control the information flow, the information provided to you have been limited, and tuned toward benefits the providers, and very soon at the expense of yours.

Smart engineering can benefit the users if it is operated with principles and ethical underlying. Most of the systems of recommendation started out this way. They are their to improve customer experience, to save time, to provide important information for decision making. Smart algorithm can be used to prevent bad trends such as someone has been searching for negative words, looking for extreme means. Ideally with some degree of certainty, a pop up windows provide an number to call if you are depressed or needed medical attention, or alert the authority if some physical harms are probable to others. Ideally - and that is wishful thinking, and often one algorithm can be certain in one way to another. And yes, technology is fast, and the response from government to provide public service is slower leaving the gap for other institutions to figure out what to do if we have such information.

I used the adjective smart as a common sense phrase. If an algorithm, device, or mean provides a conventional service in a better way, in combination with engineering improvements. We labeled them as "smart" for the time being. Smart is in the same category as intelligent or wise but more defensible use them with human's product.

So what we go from here?

I'm not sure, or at least  I do not have a concrete thought about it.

I am been worry about critical thinking of the youth. Not that I don't have a job, or have my own problem to solve, but my future, family and the place I live are dependent on how the next generation turns out. And not be that far, I need to know if that rumor makes sense to by a toilet paper or that "intel" is worth to invest in real estate, or a "spread" of information causes everyone anxiety.

Critical thinking is not some magic that gets you out of a maze nor it is the knowledge posses by a few. It is a skill set to feather out what is worthiness, more trusted vs. those deems to be fault, wrong or misled. In a sense, that your filtering to tune your database into the righteous information. It can be learned, trained, improved, and re-learned.

If you have a firmly belief on something, ask the alternative possibly. In this case, don't look for the evidence to support your belief, ask if there are different views on this. Make sure you have comprehensive, and exhausted search on the topic rather a trove of biased facts. Biased is not wrong, but the prevalence of one sides make the other side does not existed. Your thinking is hopefully balanced, not diving.

If you see many contradicting perspective, limit your search to those consider as the top trusted sources in each point of view, can be for or against, liberal or conservative, capitalism or socialism. Not everything has to be settled down to one. We are still debating many topics, even with topic has settled to one country and not others. Don't draw a hasty conclusion when you can defense that perspective in three Whs questions. In problem-solving, we have 5 Whs approaches, when we believe that by asking a series of Wh question, we can scratch the surface, grasp the essence, pull out what is essential.

Life is moving forward. What is right today, could be proved wrong tomorrow. The information presented to you has its expired date, and the scope when becomes less defensible. Know your limits is the phrase for that. Be open-minded what to come even though it is not what you think (or assume). Work on information as the pattern not a single fact or detail or a number. Learn how to think, not what to think.

US Presidential 2016 Election made history with Facebook, Qandon, 4chan. And it is just a start of what we can do with information and customized delivery on the information. US is the first test bed. Let see what we can learn or really, observe in 2020 US Presidential Election.