I have been an introvert my whole life. By Introvert, I don’t mean that I am silent. You ask me about something I know, I will talk till you are tired, strangely, I can do this with strangers too. But I struggle with small talk. Wishing someone Good morning or saying thank you or saying you are welcome are very difficult for me. One of the reasons being I think these words are pointless. I don’t help someone for a ‘Thank you’, I do it because I can. So, I expect any one able to help me in a situation to do the same. Likewise, I don’t know what kind of a day someone is having. How can I wish their morning to be good? Anyways, because I talk a lot people think I am an extrovert. But they get it wrong. It has nothing to with how much people share or talk. An extrovert is someone who seeks company of people because they need external simulation to their brains. My mind most of the time needs a break from over stimulation. That’s what introverts are. That is why I prefer solitude to company. So, I prefer minimizing external triggers. Thanks to a world wide pandemic enforced work from home way of life, things have been good. But lately I am getting annoyed by another side effect of the pandemic. The fall in revenue for big corporations..

I do not have any financial interest in the welfare of these companies. I actually think abolishing all Stock markets will do the world a great deal of good. If there is no pressure from investors to deliver Year on year and Quarter on quarter improvement in profits, there would be less corruption, less unethical practices, less targets, less up-selling, less cross-selling, less marketing and finally less marketing calls. It’s one of the things I hate in the world. Some smart-ass with an MBA reading a couple of books by so called successful people and forcing their text book non-sense on the world through desperate people trying to make a honest living(so called telemarketers)

This is exactly the corporate world right now. Its called machine learning and artificial intelligence in marketing to enhance the lead to sales ratio. In plain English it is trying to sell stuff to people based on decisions taken by a software written by a bunch of geeks with zero social skills and no understanding of human beings. This is why Google and Facebook got a heart attack when Apple pushed an update that let users decide which apps can track them. The Facebook and Google business model is to predict things you are likely to engage with, show it on your screen and show an ad now and then. Why do you think the streaming services provide different profiles for an account. That’s because they want to separate the recommendations per profile to increase the chances of you spending more time looking at the screen. Why can’t a person have diverse interests? Why can’t someone appreciate both Les Miserables and Fast & the Furious?

Its called targeted marketing and for that they need data. Data on what you like, what you bought, how much you earn, where do you live. Silicon valley was once a simple place, people wrote software and they sold it. Now they are giving it away for free so that they can sell ads.

I digress. So, the annoying telemarketing phone calls have increased a lot. I do understand that the country I live in has more serious problems than to think about making laws for privacy and customer data protection. I don’t mind getting calls for products I never subscribed for, I am most annoyed with the ones that I am a customer. Because, in this case, I should have say in opting in or out of promotions. It took me 4 months, 3 complaints, 2 escalations and 2 consumer affairs complaints to stop my bank from calling me everyday asking me if I wanted a loan. Why does this happen? Its because the stupid software are designed in such a way.

The angry face on top is a word cloud of all words longer than 4 characters from my first 53 blogs combined, minus the words that have occurred fewer than 5 times. The size of the font is directly proportional to the frequency of occurrence. It took me 90 odd minutes with the help of another blog Text Mining: Extracting and Analyzing to create it. Since you cannot eyeball the image for frequency, I made a chart below that shows top 15 occurring words. God knows what people are doing with their machine learning trainings and certiications.

If I would feed this to an algorithm to take a decision on my ideas or even better to lets say come up with a list of topics that I would prefer to read. This is something some websites do for click bait. Thankfully, Most of them do not spend time on analytics, they just try to use some low hanging fruits like donations to save a life, easy ways to earn money, weight loss tips. Or for men, its just random articles with photographs of women.

If you look at the analysis of my blog and decide that news on Music artists would interest me, you would be wrong. All 50 occurrences are from a single post about my taste in Music. Words ranked 2 to 13 can’t be used to judge be. That leaves Movies and Earth. Film gossip doesn’t interest me. I like movies a lot, so I do not watch many. That sounds like a contradiction but its not. If I like something very much, I would prefer quality over quantity. What would you sell me with earth? Donate to Save the earth campaign? If you actually read my blogs, you would understand that I believe that the there is no way humans can save the earth. The thing I most frequently read is Football news. I did write a blog on it. But machines can count words, not measure emotion or passion. I know guys in the IBM Watson team would disagree. I would truly love to see its results. I remember using it to try and measure the mood of customers based on emails. This as well before the Brisbane city council thought it sarcasm. May be it now knows Australian sarcasm. Again, I digress.

Some data analysts may point out that there is more to data science than just top 15 words. I know that. Linear Regression, Linear Regression, K-Nearest Neighbors, Support Vector Machines, Naive Bayes, Decision trees. I have learnt them all and tried them out and ultimately came to a conclusion that brains are far too complex to be simulated especially when you want to use it for selling stuff. Yes, I was only telling partial truth earlier when I said it took me 90 minutes to draw a word cloud. The time was to learn to scrape the data from a WordPress site without using paid extensions, the ability to understand and interpret data takes years.

My Amazon recommendations are so inaccurate that I turned off the notifications. May be it’s also because since the app Wunderlist was purchased and killed by Microsoft, me and my wife have been using Amazon app for shared shopping list. That may be messing up the algorithm. My Netflix recommendations have the same accuracy of my daily horoscope no matter how many movies I rate. The machine may decide the recommendations, but its a human who is tagging and categorizing the movie. Yesterday I saw a category, “Asian movies with strong female lead”. I mean what the fuck is that? My wife watches a lot of Korean movies and may be a couple of the last ones had a female oriented plot. But WTF? That’s what algorithms do. In case of movies or books, doesn’t sticking to same genre or style bore you?

Cathy O’Neil starts her book Weapons of Math Destruction with an example of a horrible failed idea of using AI to grade the performance of teachers. Read it, it’s a nice book. In the theoretical world of computers, the parameters are fixed, in real world they aren’t. The real world doesn’t abide by the rules of the digital world.

If you think that I am against technology, I am anything but that. Just like movies I like technology very much. Just as I hate bad and mediocre movies, I hate seeing technology being used for stupid things. I will give you an example, I once spoke with a guy who told me that they put GPS trackers on their construction equipment to know if they are being used on non-working days. I asked him about the accuracy of the reading. He told me 200 meters. Then I asked him “A crane or a poclain or a fork lift can be used in a construction site without moving for more than 50 meters. The only time they move more than 200 mts is when they are moved to a different site”. He did not give me a response for that.

You can use data or technology for good things too. I would prefer technology being used to eliminate malaria, not selling burgers. If people need it, they will buy it. If you make stuff that people don’t need, you have to find stupid ways to sell it. The irony of all this is that, they are asking computer programs to think like humans and asking their humans to stick to per-determined responses like a computer program .

 "Would you like fries with that?"