Meet Jacob Anderson | Intellectual Entrepeneur

We had the good fortune of connecting with Jacob Anderson and we’ve shared our conversation below.

Hi Jacob, we’d love to start by asking you about lessons learned. Is there a lesson you can share with us?
Don’t stop. My grandfather told me, when I was a kid, “Jacob, never be satisfied.” My grandma scolded him for that comment, but he was correct. You have to keep moving forward, constantly adjusting yourself and your objectives to fulfill your ultimate plan. You may not have a plan, but that’s only because you haven’t written it down. That plan, which is in your head, guides every decision you make.

Let’s talk shop? Tell us more about your career, what can you share with our community?
It’s never easy, despite what anyone at the top ever proclaims. There are trolls at every bridge, so you have to be strong enough to ignore their shade and keep moving forward. I knew what I wanted to do at an early age and pursued it when the opportunity presented itself.

Computer programming was very easy for me, so I did that as a hobby until I sold my first title to the Veteran’s Administration. That was a clinical leisure assessment tool, which is just a fancy database to keep track of a bunch of vet’s who are prescribed socialization and game play, such as billiards, shuffle puck, etc.

Science was equally easy for me, and I loved doing it. When I was a teenager I designed an electric car that looked a lot like a Tesla today. This was in the 80s and the technology for an electric car was not advanced enough. It was a great hobby to build the control software for the car, and design the theoretical aspects of the power distribution modules, voltage control for speed control, etc. I found books on robotics, electronics, electrical engineering, and mechanical engineering, and read them to figure out how to make the car work.

Most of the software I wrote as a kid was for making characters, dungeons, and supporting my hobby of playing Dungeons and Dragons. I spent days writing all sorts of data management tools that created dungeons, characters, and were lookup tools for the many books associated with D&D. Nothing was portable in the 80s, so I couldn’t take my programs with me when I would go to friends’ houses to play D&D.

I also wrote a lot of Star Trek games. These were simple games where you had a space “grid” of the turn based game. Each turn you would move the Enterprise and often would reveal a Klingon warship lurking behind some planet or nebula.

Careerwise, I started at Los Alamos National Laboratory. That was a goal, and I made that goal. My college, Worcester Polytechnic Institute, requires an MQP which is a Major Qualifying Project, much like an engineering capstone project. I was doing both computer science and nuclear engineering at the time, so I needed an MQP that would cover both majors. Los Alamos picked me up in the Science and Engineering Research Semester (SERS) program and I went to work in X-Division running tests against MCNP. Dr. John Hendricks was my mentor at the time, but he didn’t like me as a student. I was very “uppity” in his mind, and I told my professor at WPI that the work I was doing was not MQP quality work. John didn’t like that, but he tolerated me moving to something different, which was something I came up with. This was a career defining moment for me because the project I wanted to do had never been done, and John Hendricks was very vocal with me about how arrogant I was to presume that I could do something in 9 weeks that 10 staff members could not pull off in 7 years. When I was done with SERS, I got a privileged audience in the administration building where I presented my project to a standing room only audience of hairy dice wielders (that’s PhD physicists and engineers). I pulled off the impossible and laid the groundwork for my career as a staff member in X when I graduated from WPI. That work went on to be a finalist in the R&D 100 competition in 1995, titled AVATAR – Automatic Variance and Time of Analysis Reduction.

This is a common theme in my career. People tell me “that can’t be done,” and then I go and do it just to show them that anything can be done if you just keep moving forward.

I was the project lead on FOX NFL Game Tracker 2000, and then went on to help insurance companies quickly move their stuffy old practices to the internet in a secure and safe-for-insureds manner. That meant showing these companies how to do cryptography and why it was important to implement it early on in all of the databases they were keeping. My clients were decades ahead of the regulation that has become ubiquitous today.

Around 2013 I started making games again. Mostly I was bored with the kind of work that I was doing and wanted to go back to my roots. This took me into the Cocos2D-X project where I started to convert that framework over to C#/.NET. Another developer in Uzbekistan was doing the same and we merged our efforts to create Cocos2D-XNA. With that framework I launched Totally Evil Entertainment which focused on making games that highlighted what could be done with the Cocos2D-XNA framework. I created several games in the couple years that Totally Evil existed. These games were titled Santa Shooter, Mayan Tiles, and Mayan Epic. You could have played these on any device, literally. I got the framework working on every device I could get ahold of. This was thanks in part to MonoGame, which was another project I was contributing to. These frameworks, Cocos2D-XNA and MonoGame, are both open source. Sadly, some unfortunate interactions with some Xamarin people lead me to abandon work in open source, and so I closed the chapter on Totally Evil Entertainment.

Just before the pandemic hit, I enrolled in Georgia Tech’s Masters of Computer Science/Machine Learning program. Before tech, I was doing the online Coursera courses and learning about recent ML/AI advances. That was fun, but they were too easy. Going back to school to get a Master’s in Computer Science at 47 years old was interesting. That first ML class with Charles Isbell was a stark awakening to how soft I had become. I didn’t do well in that first ML class, but I corrected my attitude and got back on track. My capstone moment at tech was back in Isbell’s class CS7642 – Reinforcement Learning. It was in that class that solved the Lunar Lander problem using a classical double-Q learner. This had never been done before, and nearly everyone said that it could not be done. They always said “the state space is too big,” or “you have to use a deep Q learner.” I didn’t believe it, and there were times of doubt, but I kept at it and it worked. As far as I know, I am the only person to ever solve Lunar Lander using a double Q learner. No neural network was involved in that solution, just straight up ordinary Q learning.

Then the pandemic hit and the XPrize held the Pandemic Challenge. Why not, right? This was another chance to solve a big problem using the AI/ML education I just finished. This problem was quite a bit harder to solve since it involved getting data from a variety of sources, and adding my own data sources that were relevant to the pandemic. I made it to the final of the XPrize and took my predictor model public. You can see it now on the Beyond Ordinary Knowledge page. This was a fun project to automate, both the training side and the predictor side. Every day it updates the predictor rollout without any involvement on my part.

If you had a friend visiting you, what are some of the local spots you’d want to take them around to?
We’d go to the Cabrillo monument, which is the lighthouse on Point Loma. That’s a nice view of the city and a good view of the ocean. Then we’d drive over to Coronado across the bridge since that’s an iconic landmark for San Diego. The beaches of Silverstrand are very nice and easy to access, so that’s the best beach experience in my opinion.

The zoo, the animal park, Sea World, all of the attractions for a day each. The zoo at night in the summer is weird because of the peacocks. Listening to those nutty birds squawk all night is worth a visit. The roar-and-snore experience of the animal park is a lot of fun too, listening to the lions moan at night, and the elephants talk to each other.

As for food, I would take them to Milles Fleurs in Rancho Santa Fe. The food and service there is worth the price. That’s a capstone evening meal, though, so save that for the last night.

If this was in January or February, then we would go out on the ocean to see whales. A small sailing boat experience would be best so that we could get closer to the whales. Plus sailing on the ocean is a treat.

Who else deserves some credit and recognition?
Vernon Hagedorn, principal of Hot Springs Middle School. Vern is one of those people who could see beyond the horizon. If it wasn’t for Vern then I would not have had the early exposure to computing devices, and that would not have given me the opportunity to realize my calling to software development. Then my late uncle Scott Clark who taught me about business and people, who was a serial entrepreneur himself and inspired me to take on my own business. Darwin Schroeder my high school computer science teacher who adopted early computing as a curriculum and exposed me to more advanced programming methods, plus he took us to programming contests at the local colleges.

Website: www.beyond-ordinary.com

Image Credits
Emily So was the artist who did contract work for Totally Evil Entertainment.