Self Driving Cars

A dilemma about how to program a self-driving car showed me how interdisciplinary some problems can be.

Today was my first STEM Skills and STEM in Society class. The class started with a TED-Ed video that raised an interesting question: If you were programming a self-driving car and a box fell out of a truck, would you program the car to let the box hit, swerve into an SUV, or hit a motorcycle. Here is the link to the video. Here's a picture of the situation:



This is a hard dilemma to answer. If I were the programmer, I would research each situation and find or calculate the average number of deaths for each situation. Then I would choose the one with the least number of average deaths. This method allows the choice to be made purely on empirical data and not affected by human biases. For example, if the program had a bad experience with a motorcyclist, he would likely choose the motorcycle. Our biases affect us whether we are conscious of it or not.

This problem is interesting to think about. It brings up many philosophical questions like

  • How do we value life?
  • Who is to blame? 
  • What principle or theory should be applied to help solve it?
and many more. The many unknowns also make this problem difficult to answer. This exercise taught me that there are many places where different fields intersect. However, critical thinking and reasoning skills can be used in these places to find an answer.




Comments

  1. Kristina GehrmanJune 13, 2024 at 8:40 AM

    I like the way you considered the possible bias of the human programmer! It foreshadows the conversations we've had on objectivity and values in science and shows us how these same issues arise in tech and applied contexts.

    ReplyDelete

Post a Comment

Popular posts from this blog

My Gov School Experience

Hello, my name is Luke Akers. I am a rising senior from Memphis University School. At school I am involved in many clubs and play clarinet in the band. In my free time I like to hang out with friends, listen to music, and play video games. My favorite school subject is math which led me to Governers school. I have always been good at math and enjoy it. It is like a difficult puzzle you have to solve. I have some background in the subjects taught at the math program such as matrices; however, I hope to gain a deeper understanding of these concepts. I also hope it can help me for Advanced Topics, a challenging math class I am taking at my school next year. I hope GSSE can help me improve my STEM skills and meet like-minded people. I also hope it can help me narrow down a career choice and gain exposure to the college lifestyle. 
Read more

June 13: Are Animals People?

Image
 Are Animals People?  Today in STEM in Society we discussed personhood. Are aliens persons? Could any animals be people?  If you do not meet all the criteria for personhood, does that mean you are not a person? I believe there are different levels of personhood. A human has a different level of person hood than a elephant, fish, or rock. However, intellegent animals still deserve some basic rights. For example, Happy the elephant has lived in solitary captivity for 16 years! This is inhuman as elephants are intelligent creature and she could easily be moved to a elephant sanctuary. Here's an article about elephant's intellect: https://wildlifesos.org/chronological-news/the-remarkable-cognitive-abilities-of-elephants/ Here's a picture of Happy:  On the other hand, fish have little to no personhood and rock definitely have none. We do not even know if fish have consciousness and rocks are not alive. 
Read more

June 17: Data Visualization and AI Overview

Image
Data Visualization  Today in STEM Skill we learned about data visualization. We followed a YouTube tutorial for Excel and made a report on the sales of a product in South America. I learned how to create and edit graphs based on sales data. Specifically, I learned how to manipulate the graphs to show data effectively. AI Overview Artificial intelligence (AI) is a computer learning model that learns based on the data given. It generates responses to questions asked by users based on this data. My team used gender, class on the ship, age, and our premonitions about each character to predict who would survive. Gender was our main predictor because, in the movie clips, only women and children were allowed on the lifeboats. We also used class since one man bribed the captain to get a lifeboat and first class was higher on the ship (they would have more time to get on a boat).  The code had a 0.77511 success rate which is higher than using just gender. It did well overall. Through this exerc
Read more