So How do I get to General Motors?

While on the adjunct hustle, my partner at the time was working for IBM and we got a call one day. We were living in the DFW area and they wanted everyone on their team to relocate to Austin, TX. The question was asked as “You’re planning to move to Austin, right?” and so given that they were the primary bread winner we planned to move to Austin.

While looking for jobs I came upon a listing for an “Artificial Intelligence and Machine Learning Scientist.” At this point I had some machine learning experience doing non-parametric Bayesian Learning before grad school, so I decided to apply. While I was driving to the office to teach the next day I got a call from the recruiter saying that the hiring manager wanted to talk to me immediately and so could we schedule that call. Later that day we talked and eight days later I had an offer from General Motors to join their AI/ML IT Center of Excellence.

In my first role at General Motors I was given significant latitude. I solved problems using AI, ML, stochastic differential equations, topological data analysis, diagrammatic calculus for braided monoidal categories, and everything else that appealed to me.

In the process of all of this I learned a few things. In graduate school, my only criteria on whether I should work on a problem was whether or not I found it interesting. I had a good set of leaders who took several tack to this. Some would present me with a problem and then sit back and let me convince leaders to solve this related problem that was much harder and also theoretically more fulfilling. Others taught me how to change the way I evaluate problems. Interesting is just one criteria, but there are other criteria out there worth considering. An example of that is the value provided by the problem. Now, value can mean a lot of things. One of those is how much money the problem is worth, but other is what it provides and how it works with other initiatives. I learned that there are just as many interesting valuable problems as just interesting problems (fun with infinities.)

Eventually I found that the problems I was putting together to work on weren’t just things I could do on my own. I needed teams and started becoming a technical leader. Please excuse the hyperbole, but I learned that everyone wants to be Michael Jordan on the field and make the game winning shot, but there are others. I ended up as a hybrid of Magic Johnson and Dennis Rodman. Magic Johnson because I wanted to be the one who controlled the ball and passed to whoever made that game winning shot. Rodman because I realized that when putting teams together various skills play various roles and gaps are naturally going to arise. I became an expert in identifying those gaps and then becoming sufficiently competent for the project to proceed until someone better came along.

When the guy who hired me into the CoE left, I raised my hand and said that I wanted to be the person who lead the team. I became a manager. This brought its own set of lessons, but generally I find those off topic. There were ways in which I was good at it and ways in which I wasn’t. It helped that I started out in an environment where I had built trust and was working with people I knew well towards common goals. Over time though, that group got dissolved a few times and the work changed.

Back to Being a Researcher

Luckily for me there was a coup. During one of the last rounds of re-orgs I proposed that it was time for a new scientist to learn the skills of being able to manage people. There wasn’t much more that I was going to get and the work had moved away from solving hard problems. So I was already back to being an individual contributor and getting my legs back underneath me. To quote Jay-Z in Encore:

“When I come back like Jordan
Wearin’ the 45, it ain’t to play games with you
It’s to aim at you. Probably maim you
If I owe you I’m blowin’ you to smithereens”

And that’s what it was about. I wanted to get back to producing intellectual property and patents and get back to doing the technical portion of solving problems.

Then, in April 2026, the coup happened. I got scooped up into Research and Development into a new organization focused on emerging technologies and artificial intelligence research. We were to focus on applying advanced methods in these fields, Now I’m working on how to apply artificial intelligence and machine learning together to solve monumental problems that industries are based around. Is it fun? Oh yeah. Can I say much more about it? Not in the slightest. But it’s pretty awesome.