A Career in Academia

Academic Research Summary

In a separate blog, I will summarize my academic journey so those of you who want to either know what that journey might look like, or would like to put these comments into context can read about my experiences.

Academic Value System

Value systems are very important. It will tell you what will be rewarded and if your passion and/or skill strengths align with that value system you will be rewarded. I want to be clear that these conclusions are based upon my experiences and the experiences of others may lead to different conclusions. However, I will say that I have talked to a lot of people who are in academia and industry and for the most part these views are shared.

Teaching vs. Research

My journey led me through a number of high-powered research institutions and as such there is an emphasis on research. Teaching in these institutions are neither a focus nor is it truly valued. I was told explicitly when I was a professor that as long as you are getting ‘3s’ on your teaching ratings, you’ll be fine and if you your teaching is not the greatest, do not worry about it. Personally, I enjoyed teaching, but I fully recognized the value system that I was in at the University. So, I would always tell my students “I will make all of my notes available on-line. You will have all of the readings. If you don’t find any value in my teaching, don’t come. I will be paid the same either way.” I also would tell them, “You are attending a high-power research institution. If you are only going to classes and not taking advantage of working in a research lab or with a professor you are not getting the full value out of your education.”
If your passion is teaching, go to a teaching college. I have some good friends at Carleton College who came from high power research universities and decided to focus on teaching. They are extremely happy with that decision. If you decide to go to a high-power research institution, your path to success will be through research papers and grants.

Research in Academia

Research in academia is great. There are an infinite number of questions that you can ask and sometimes it seems that professors are doing just that…asking an infinite number of arbitrary questions (IMHO). However, my experience was that when I finally made it to be a professor, I actually spent less time doing research than I did as a graduate student or a postdoc.
When I was a professor I spent my time in the following way.

  • 30% Teaching and teaching prep (~4 courses a year; 1 graduate, 1 upper-level undergrad and 1 lower-level course). Even when I had my lectures and materials completed, I still had to spend time in the classroom, meeting with teaching assistants (TAs), students, preparing exams, etc.
  • 20% mentoring undergraduate/graduate students (I enjoyed this part of my job. I did more than I was encouraged to do.)
  • 20% writing research grants (I was an assistant professor looking to get funding for my lab).
  • 10% writing research papers (I did not enjoy this part of my job so I didn’t spend as much time doing it).
  • 5% department and university ‘church work’ (meetings, admissions, teaching reviews, etc.)
  • 5% Research: coding, designing experiments, reading research papers
When you become a professor at a research institution you don’t spend your time doing what got you there (research) but end up spending most of your time managing and building your lab. In my experience it is much like creating a startup in the safety of a company where your paycheck is guaranteed.

Academic Career Path

  • 4-5 Years Undergraduate (age 18-23)
    • Income: 0-$25,000 (campus/service jobs)
  • 4-6 Years Graduate Work (age 23-29)
    • Median Income: $28,965 (payscale.com)
    • Supported through teaching
      • 10-25 hours of teaching/week
      • ~25 hours of lab research
    • Research Grant Support
      • 40-60 hours of research / week
  • 1-4 years of postdoc (age 29-33)
    • 9-month Median Income: $48,455 (payscale.com)
    • 40-60 hours of research / week
  • ~ 6 years Assistant Professor (age 33-39)
  • ~ 5-10 years Associate Professor [Tenure] (age 39-49)
  • Full Professor [Tenure] (49-retirement)


Career in Academia vs. Industry

https://commons.wikimedia.org/wiki/File:Researcher-test.jpg

Summary

I have had the good fortune of spending more than a decade in academia (graduate student, post doctoral and professor) in addition to more than a decade in industry at a fortune 100, multi-national corporation. I have often been asked by interns, graduate students, post docs, ‘Why did you move from academia to industry?’ followed closely by ‘What is the difference between the academia vs. industry?’ I decided to take some time and write down my thoughts on the topic. Below you will find my thoughts and perspective.

Framing the question

Academia and industry both can provide very fulfilling careers. This decision (like most life decisions) comes down to an HONEST assessment of one’s set of values. Before beginning the discussion I always try to make the following three points VERY CLEAR:

  • There is no ‘right answer’…just different paths. If you are asking this question you have been part of academia for almost your entire life with tests and research projects and in that process you have been lead to believe that all questions have a ‘right answer’ and you worked very hard to find those answers. However, for this question there is no right answer, there are just different ‘paths’ and on both paths can lead to fulfillment, happiness and success if you let them (it is up to you).
  • The decision is not permanent. As I’ve experienced my career and watched other careers (and this is generally true of life), straight lines are drawn with crooked paths. Moving between academia and industry can be very difficult (both directions) but not impossible.
  • You didn’t get here by making bad choices. Whether you have a Ph.D. or you are considering getting a Ph.D., your accomplishments are built on a series of very good choices. Therefore, in this case all of the ‘bad’ options are off of the table (e.g., ‘post doc or become a crack addict?’). If you are having a difficult time deciding then have confidence that both choices are excellent choices but that they are just different (see item #1 ‘no right answer’).

So with that said, both academia and industry can provide excellent opportunities for a fulfilling, challenging and rewarding career.
opportunities for teaching, research, working with smart and talented people, wealth and making a difference in the world. However, being able to do these things in academia or industry can either be ‘natural’ or ‘forced’. When ‘forced’ that means it is possible, but you will have to bring your organization outside of their comfort zone to make it happen. When ‘natural’ it is the natural part of your career path.

Tutorial: Robot Operating System (ROS) on AWS-EC2

I spent a Sunday morning figuring out how to run Robot Operating System on AWS EC2. Here are the basic steps and the references to get it working. I’ve gone through these steps probably about 10 times today and it takes about 5 minutes to set up an instance and configure it.

One of my weekend challenges is to understand how to use Docker better. I might make a “container” or “Docker Image” for ROS. If I do, I will put the reference here.

Installation

  1. Install Ubuntu-16.04 AMI (current stable Ubuntu version)
    Ubuntu AMI
  2. Follow the instructions on how to configure your Ubuntu Instance for Desktop Access
  3. Install ROS Kintetic
    1. Do the ROS Installation using the remote desktop.
  4. Work through ROS tutorials
    1. Do the ROS tutorials using the remote desktop.
    2. Here is the first ROS Tutorial in the series that has something that will display.
      ROS on AWS EC2

Issues

  • I had an issue with roscore. However, I was able to do the rest of the tutorials.
  • I tried to install and run the ROS Turtlebot but I couldn’t get it running. Here is what I found out.
    • Turtlebot requires ROS Indigo
    • ROS Indigo requires Ubuntu Tasty. However, I couldn’t get Ubuntu-Trusty, ROS-Indigo, and Remote Desktop to all work together.

Tutorial: Connecting to AWS EC2 Ubuntu Desktop

I spent a Saturday morning trying to connect to an AWS EC2 desktop after trying to follow the instructions given by AWS on Connecting to Ubuntu Desktop with Windows that completely failed.

The instructions described here are derived from the following YouTube Video. There were some parts that were not clear/correct so I have clarified and corrected them in this post.

Configure Ubuntu Desktop Server

sudo -s
  • Update the operating system
sudo apt update && sudo apt upgrade
  • Type the following commands to install vncserver:
sudo apt-get install ubuntu-desktop
sudo apt-get install vnc4server
sudo apt-get install gnome-panel
  • Start the VNCServer.
vncserver
  • Remember the password you set. You will need this passcode later when you connect using tightvnc .
  • Kill vncserver
vncserver -kill :1
  • Modify the VNC startup script
vi .vnc/xstartup
  • Add the following lines to the .vnc/xstartup script
#!/bin/sh
# Uncomment the following two lines for normal desktop:
unset SESSION_MANAGER
# exec /etc/X11/xinit/xinitrc
gnome-session –session=gnome-classic &
gnome-panel&

xstartup script

  • Press ESC, followed by :wq (write-quit) to save and exit the file
  • Start the VNCServer again
vncserver

Configure your Windows Box

Configure your EC2 Instance

  • Edit your security group.
    Edit your Security Group
  • Add port number 5901 in your ec2 security group
    Add Port 5901 to security group
  • Write your public ip number (or DNS name) in remote host text box and port number <publicIp>::5901
    TightVNC
  • Your desktop in ec2 instance is ready and execute the command vncserver after every restart.

FAQ

  • Can I stop my machine and restart it?: Yes. However you will have to re-run the vncserver command after restarting the server.

Ready Player One: Not a “Star Wars” Moment

I went to “Ready Player One” last night with my 12-year old son looking for a “Star Wars” moment with him. He and I had read and listened to the book and we were excited to see the movie. The references in the movie were to many of the things that I grew up with as a nerd/geek. Dungeons & Dragons (D&D), Atari systems, quarter fed gaming systems, hanging out at pizza joints, etc. There was also the high-tech rigs, Gundams, and enough action and heroism in the book to make any preteen excited. However, Spielberg was extremely disappointing in this movie. I have a hard time believing that he “actually” directed this movie and didn’t simply sign his name to the movie and directed it in abstention.

I know that a lot of people complain about directors taking liberty with the story line of a book, so at some level this rant might be an old rant. But I just went to the movie “Ready Player One” with my 12-year old son and even he was disappointed in the movie.

Walking out of the movie I asked him “So, what did you think?”

“I don’t know.” which is his typical first response when he wants to think about the question. Knowing this, I just waited to hear what he had to say before providing my opinion. “Where were the gates? I mean, that was a big part of the book. And that wasn’t how Parzival found the first key. What was that about? And Art3mis and Parzival don’t meet in real lif (IRL) until the end of the book. Why did they rush that?”

The onslaught of inconsistencies between the movie and the book just came coming. Given that he is 12, he almost immediately communicated to his friends that he saw the movie. Later that evening as we were sitting down for his “second dinner” (he is growing like a weed) he said, “I thought maybe if I hadn’t read the book that I might have thought that it was a good movie, but a friend of mine who didn’t read the book also thought that it was a bad movie. He thought that the love story was rushed …” (which it was) “and that the story line was rushed”. Which it also was.

I remember a similar thing happening with my daughter (who is now 17) when we watched the movie “James and the Giant Peach” after reading the book. However, the complaint was much different. She was a bit younger (probably closer to 8) but her complaint was that the movie was missing her favorite scene from the book. “Dad, where were the cloud giants? That was the best part of the book.” At that time I explained to my daughter that it can be difficult to “cram” all of the items from a book into a movie and so there can be a need to remove some of the scenes. Although I agreed that there were a lot of other scenes that could have went, sometimes these things are a matter of opinion and the director gets/needs to make some difficult decisions. However, removing some scenes is VERY different than effectively changing the entire story line. For example, one of my major gripes was that a major part of the book was that Parzival creates a plan to take down the shield by becoming an indentured with the IOI. In the book, this was a major part of the story line. Parzival was willing to sacrifice himself for the Oasis. A bit of a “Christ” reference for the storyline. However, in the movie, Art3mis is the one who ends up in the IOI. “What!!??”

Spielberg has been an amazing director and has influenced my life a great deal through his movies. However, I wonder if he directed this movie in abstention. That Spielberg simply put his name on the movie but really didn’t put his mind or soul into it. If so, it is disappointing. I was looking forward to having a “Star Wars” moment with my son and this movie completely missed it. Although my son and I weren’t able to have a “Star Wars” movie moment, we sure did enjoy experiencing the story line of the book.

Cognitive Prosthetics & Cognitive Shadows

The other day a colleague and good friend of mine told me that he took all of the texts that he and his father wrote back and forth over the years and wrote a machine learning algorithm that when queried would return a text that “sounded” like his father. He said that it wasn’t exactly like his dad, but it did have a resemblance. This made me think about the concept of cognitive prosthetics in a much different light than I had before. Specifically, I realized that our digital breadcrumbs will ultimately lead to a cognitive shadow that will impact reality beyond our physical lives.

Cognitive Prosthetics

When I was in graduate school (in the early 1990s) a friend of mine and I were talking over a few drinks and we were discussing  Personal Digital Assistants (PDAs). In this somewhat altered state of mind we realized that these were not just devices but that they were actually cognitive prosthetics. When I was a child, I remember when my mind was “cluttered” with information that required such precision that one small error would make the information meaningless. During that time most people simply “remembered” everyone’s phone number (or at least the numbers for your good friends and family). Obviously a simple transpose of two digits or mis-remembering one digit would make the information completely meaningless. I had to remember phone numbers and addresses and calendar events. This seemed to clutter my mind with things that it just was not designed to encode.

However, this was not that big of a step over simple pen and paper that was used by most adults at the time. Black address books and pocket calendars were very common. However, there was something different that happened when these devices came out. When I started using my first Palm Pilot I could “feel” my mind become less cluttered. Or I should say, I could “feel” my mind able to relax slightly and focus on what it was designed to do–understand, imagine and infer.

According to a recent Pew Research Center Study 77% of Americans own a smart phone today. According to eMarketer, Americans now spend over four hours on their mobile devices each day and according to MediaKix we spend almost two hours on the top five social media sites (FaceBook, YouTube, SnapChat, Instagram and Twitter). These statistics show how much we have decided to “offload” to these cognitive devices. Initially it was simply offloading factoids that our brains were really not designed to encode (i.e., precise digit memorization is not what we are good at). We have now started to offload our social functions too. Initially it started with Facebook and efficiently “updating” our friends with our “status changes” with posting pictures of the new dog, our children or our vacations. However, this now has grown even further. YouTube provides a method for efficiently transmitting knowledge and teaching to millions of people immediately. Instagram and Snapchat allow us to rapidly communicate images of where we are and how we look and feel.

Cognitive Shadows

I am approaching my 50th birthday this year and as I look back on my life I think about what an amazing period I have lived in. When I was in high school I still typed my papers on a manual typewriter. Today, I hold in my pocket the knowledge of all of humanity. As I look forward to the next ~50 years I think about what might change in those 50 years. Cognitive prosthetics are about remembering what happened in the past. But what if all of those digital trails that I left behind could be used to create a Cognitive Shadow. That is, like my friend who turned the data from his text exchanges with his father into a “text bot” that would respond with words that sounded like his father, it doesn’t seem to far fetched to think that my digital trail can be used to create a digital bot that will extend beyond my physical life.

It isn’t that far fetched to build an email bot where people could send emails to my Cognitive Shadow and it would respond in much the same way that I would. In fact, this has already been done by Eugenia Kuyda with grief bots following the death of a good friend of hers in 2013. With the digital trail it isn’t too far fetched to even have a Cognitive Shadow that not only responds, but also initiates a conversation using the statistics of my email correspondences.

But could we have a Cognitive Shadow that goes beyond grief bots. Again, the digital trail that I (and most people reading this article) will leave behind will be quite massive and getting larger every day. We already have bots that do stock trading, and there are bots that look at my trading behavior and make recommendations. What if they simply made the trades for me after my death. My Cognitive Shadow would now have a method for generating revenue. It would also have access to all of my movie preferences on Netflix, Amazon and Google Play. It would be able to continue to “experience” movies and entertainment in much the same way that I would. My grandchildren would be able to interact with my Cognitive Shadow with contemporary information.

Summary

It isn’t too far fetched to believe that very soon almost all of us will have a Cognitive Shadow. A bot that will have the capability to make decisions and interact with others in much the same way that I would act. Although, in principle, this Cognitive Shadow could extend for centuries, it would not be me. It would be a lot like the Searle’s Chinese Room argument against a machine having consciousness. Although I would not experience this Cognitive Shadow, it would be an entity that would continue to have an impact on reality beyond my life.