Saturday, January 27, 2024

What makes a good decision?


I think this a good example of what close to two million people like to read daily. It makes you feel good, feel like there is an order to the things that you do and that if you prepare a PowerPoint presentation with the right animations illustrating these principles then your boss is going to tap you in the shoulder, good job!

But are people not able to think anymore on their own? 
Can anyone stand up anymore and say, no, this is not right, it does not make sense? 
Maybe it is better to just sit in the back of the room, sipping coffee from a mug with an inspirational quote and wait until 5pm of Friday because this weekend there is the game!

Based on followers numbers, it appears that these days workers seem to think only by slogans, heart warming quotes based on words play, and the "inspirational" automatic reply on LinkedIn and not that they would ever want to argue with me, (why spend some of my time with someone who does not have any followers?), but for the sake of the argument let's apply the proposed decision making process to problem of getting married. Let's follow the steps and the reader be advised that I am actually happily married (love you honey! ❤️):

1. Define the problem:  well, I want to get married!
Defining the problem in this way implies that 
    (a) this is a problem (ok, well at least for me now this is a problem) and 
    (b) there is a solution to this problem, i.e., that there is another individual that want to solve the same problem with me. 
But, oh, wait. Why just one individual? Perhaps a polygamic solution would be a better option, after all those who have agreed to that compact have found already a solution and they can be qualified therefore as experts. Experts do know better after all. Oh my head is spinning and I am questioning myself already if this is the right formulation of the problem. And worse, I have been trying to think logically, but that is only at step 5, I cannot do that at this stage!

2. Find Options: yes, options are good, I can only choose and therefore exercise my freedom only when I have options, else I would not be free and may need to enter into a pre-arranged marriage. Oh dear, I need to go back to step 1, I had not considered a pre-arranged marriage. Perhaps this is the solution already, maybe I am going to be lucky, and anyways the experts that have arranged this marriage know it better, they are experts after all! 
But I like options, where do I get them. Mhmm, I know Sarah, and Liz, and what was her name again? I need to evaluate fully my options: I am going to ask Sarah for a date, and Liz for a date and if I find her number I will ask (darn it, what was her name?) also for a date. Should I ask Sarah first, or perhaps Liz? and the third lady was certainly Sarah's friend, so I can ask Sarah for her name and number. 

3 Collect Data Maybe, I can invite the three of them on the same date so that I can make an unbiased experiment that does not depend on the order on which the dates were defined. Maybe if explain them my intent very well at step 1 they will understand and they will allow me multiple experiments, and counterfactuals. I will collect all the data that I can get from their social profiles, plus estimate measure of their body types, your know brassier size and all that good stuff, and don't forget the age, yes, especially the age of a woman is very important, and medical history. I will create a good data set so that I can run some machine learning models, there are so many open source models on github to choose from and see what they get me as an answer.

4. Evaluate the impact Before I pop the question, though, I need to carefully evaluate the impact, especially on my finances, maybe I should put in some expenses for a marital counselor, and the lawyers, ... ah yes, I forgot to evaluate the impact of a divorce. And not only on me, if we have children, yeah another consequence that I did not consider on step 3. I need to go back there and reconsider the machine learning models. But wait, I cannot go back to step 3, because that it is not in the flowchart on how to make a good decision. Maybe I did not use enough logic on step 3, but that was not contemplated in the flowchart, it is only required at step 5! Oh well, I have to trust the process!

5. Think Logically. oh, now you tell me to think logically? Think logically about what? I have already decided the problem, the data, and the method, what is that I need to think logically now?

6. Take Action. The most accurate seemed to be a semi-supervised deep neural network. I ran it on a AWS Kubernetes cluster with GPUs, and the results were clear. Sarah 99.8%, Liz 99.7%, and what's her name only 99.1%. I just popped the question to Sarah over email, and if she says yes I will announce our engagement on X, formerly known as Twitter. Oh, wait, she just answered, I am so excited!! ... what the F? ... she just sent me to hell? How is that even possible? I have followed the process on how to make a good decision to the letter, and could I have possibly gone wrong? I am confused .. let me take a look at what is on step 7? Define the problem? but I just did!! Oh well, I need to trust the process after all making a good decision is just doing the same thing over and over again and expecting different answers, am I right?

_______________________________________________

As Author Nassim Nicholas Taleb aptly put it, "lecturing birds how to fly"


No comments: