Looking back at 2019

As 2019 was closing I had a feeling that I had not done that much during the year. But then I started looking over my journal entries, my photographs, the books I read (mostly listened to), the progress we made at Generable, and I realized that I did get some things done and experienced joy (and sometimes agony) along the way.

I started the year by going to see The Jungle, a play that plants you in the middle of the Calais Jungle and follows the lives of refugees struggling for survival. Jacki and I came away deeply moved by the experience.

Later in January Colleen Chien invited Dan, Jim Savage, and me to give a short talk at her class at Columbia Law School to discuss facts and fiction in AI/ML ecosystem today. I mostly focused on our work with Stan in Clinical Research and how probabilistic modeling is making it possible to construct models that are explainable, transparent, and perform very well predictively, especially if we are able to approximate the data generating process well.

In early February we celebrated Andrei’s first birthday. I never thought that I would be a father again but Andrei is bringing so much joy to my life that it almost seems worth it. OK, it is worth it. I think. I am pretty sure. I love you, Andrei!

Winter came in March to our neighborhood and we had some fun times in the snow. Andrei loves being outside and I can not wait until he is old enough to go skiing with us.

In early March, Generable had our first off-site in Pocono Pines. This is when started seriously thinking about what product we wanted to build. At the time we thought we are going to make a tool for Stats and Data Science types to build models from different components. That turned out to be wrong. More on that later.

In the middle of March, my oldest son Ben and I built a computer from individual parts. He was surprisingly enthusiastic about this enterprise and I enjoyed working with him on this. When we turned it on and loaded the OS (Ubuntu) it became obvious that the project succeeded. Ben had not used this computer since and I turn it on only occasionally but the whole experience was totally worth it. Thanks, Ben for putting up with my relentless pursuit to turn you into a nerd.

Later in March, we saw another play at St Ann’s Warehouse called The B-Side. This play is a musical in an off-broadway sense of the word. The main character sings along a vinyl album containing songs by African American convicts in a Texas prison. I love seeing this kind of production making it to the serious stage and selling out a large theater in Brooklyn.

At the end of March, we went to see Marys Seacole (based on the life of Mary Seacole) at the Lincoln Center. The story follows Mary throughout her life and to the battlefield of the Crimean War. I don’t remember a lot of details from that play; perhaps it did not leave an impression on me or I just drank too much bourbon shortly thereafter.

In April, I was invited to be on a panel with other alumni of the Columbia Univesity’s MA in Statistics program. I love coming back to the stats department and talking to current students and recent graduates. I usually tell them to learn some Bayesian stats — most of them will graduate without encountering a posterior distribution. A tragic state of affairs, but that’s how it is for now.

Later that month, we started re-designing the Generable platform and focusing on what we call the Clinical Lead — the person who oversees early clinical trials and gives an opinion of whether a treatment should advance to a late-stage clinical trial. Inside a Pharma company, this is not just a clinical decision, there are economic factors at work, but the clinician makes an assessment of the drug is working. We abandoned the model-building idea and instead embraced communicating model results and supporting decision-makers.

At the end of April, we went to see Oklahoma on Broadway. I know people love this musical, but something about it did not click for me. I love that it is made and I think I understand the scope, but I could not quite grow to love it.

Northampton, MA

In early May I was visiting a colleague in Northampton where we have a small office. I was staying in an Airbnb house inhabited by an artsy old lady.

“What do you do?” she asked me one evening when I came home late, slightly drunk.

“I am a Statistician working in clinical research, early clinical trials in Oncology.”

“Did you say Oncology?”, she asked.

“Yes”, I answered.

“Thank you for everything you do! I am a cancer survivor.”

I was completely taken aback as this never happened to me before.

“I am not a doctor, I do not treat patients, and if we make any contributions, it would be many years from now”, I told her.

This was not false modesty. I really did not feel that I deserved her thanks, not yet anyway. But she wouldn’t have it. I finally told her she is very welcome and now completely sober and slightly teary-eyed stumbled upstairs and went to bed.

The Confluence Museum. Lyon, France

Later in May, I attended and co-presented at the Bayes-Pharma conference in Lyon, France. Marmaduke Woodman from the University of Aix-Marseille and I talked about the work we did fitting Stan models to epileptic seizures data collected from electrodes implanted in patients’ brains. The hope is that these models could be used to improve the precision of surgical interventions. I think they are planning a clinical trial for later this year.

Riga, Latvia

In June, I attended the PAGE (European Pharmcometrics) conference in Stockholm, and right after the conference, I caught a short flight across the Baltic Sea to visit my mother in Riga, the city of my birth. Riga is a beautiful, modern European city with manicured parks, well-maintained Art Nouveau architecture, mild weather, and tragic history.

Every year, I promise myself that I would spend some time with my parents and this year I kept my promise. We have a lot in common my mother and me, kindred spirits so to speak. For one, she is just as vulgar as me and she appreciates my not-so-kosher jokes.

On the way home, I had a stopover in Amsterdam, where I spent a few hours at the Rembrandt House Museum before taking a long flight back to New York.

In June, my daughter Miriam graduated from middle school. She worked really hard and improved her grades considerably. I was (and still am) so proud of her.

At the end of June, the Generable crew had the second off-site meeting in Denver. I like spending one week with our remote team every three months or so. It helps to get on the same page, agree on key priorities, collaborate on technical tasks, and just spend some time hanging out together.

At the beginning of August, we spent a customary week on Lake Sunapee in New Hampshire. Ben took some sailing lessons, we played some tennis, biked, played Monkey Bridge (a Buros family tradition), and generally had a nice relaxing time. I worked most of the time but it did feel like a vacation.

On the way back from Sunapee we stoped in Belchertown to break up a long drive back. In the morning, Ben and I took a 10-mile bike ride to visit the University of Massachusets at Amherst. It is a lovely suburban campus with lots of green lawns. Can you picture yourself going here, I asked Ben. I dunno, replied Ben which is his standard reply. At least he is not over-confident!

In the middle of August, the Generable crew attended StanCon in Cambridge UK, an annual conference dedicated to all things Stan. Generable was one of the sponsors and Dan and Krzysztof presented.

In September, we took Andrei to the New York Aquarium on Coney Island. They have recently renovated the place and some of the construction is still in progress.

At the end of September, Jacki, Miriam, and I went to Disney World. Miriam wanted to go for a long time and I am happy we were able to do it. The Magic Kingdom is aging and not very gracefully, but the Flight of Passage ride in the Animal Kingdom left such an impression on me that I am seriously considering getting a VR set even though the ride itself is not in VR. It’s pretty damn close to R. While in Orlando, my dad came and stayed with us for a few days, which was nice as I do not get to see him that much anymore.

In October, we showed an alpha version of the Generable platform at the ACOP conference in Orlando. This is the first time we were able to afford a booth, which is some kind of milestone. A lot of people don’t like “working the booth” but I do, particularly when the traffic is heavy which was not always the case at ACOP. Next year we should be doing more clinically oriented conferences but we will likely be back at ACOP and PAGE.

In September we went to see Slave Play, a Broadway production that is too weird to describe so I am not going to try. If you go see, and perhaps you should, it will make you very uncomfortable, which I am sure is by design.

We ended the year with the play The Sound Inside with Mary-Louise Parker (Weeds and other good stuff) in the lead role. This was my favorite play of the year and one that I will remember for a long time. Poignant references to DFW, parallels and direct references to Raskolnikov from Crime and Punishment, a masterful soliloquy by Mary-Louse, are just some of the features that made this play special for me.

This was in many ways a theatrical year.

Here is to you 2019!