Pivotal discussion at Whole Foods

At Gail’s in Barons Court station. It’s wonderful to temporarily be back in the old neighbourhood.

Last night, M and I had supper at Whole Foods in Fulham, and our conversation about the future of my work was a pivotal one. I had just read about OpenAI’s latest release – Code Interpreter – which is able to complete entire data analyst tasks. Reading these posts and related material it’s become very clear that much of what I love about my work is going to disappear: the craftsmanship involved in writing concise code, thinking about what method is most appropriate for the problem at hand, and visualising and writing up conclusions. Whatever part of that current models aren’t able to do well, future models will in due time. It seems highly likely, therefore that the work of most data scientists will turn into something more of a product manager, as they will manage different AIs to complete all the low level tasks and merely manage and overseeing those AIs. I may or may not enjoy doing this, but it surely isn’t why I moved into data science.

However, what I’ve realised over the past few months in my current job is just how scarce skills like conceptual thinking, clear communication, and accurate prioritisation are. This isn’t the case at all companies, and certainly not at top companies in their field. But overall, it is true, I think. I went into data science not least to find out whether I wanted to become a technical specialist or move into a more managerial role. What yesterday’s discussion has clarified more than any previous one is that I should see this inflection point as an answer to that question. I am pretty good at thinking conceptually, grasping what is essential in a discussion or a concept, have a very solid technical understanding, love teaching, am a strong communicator, would love to help people grow in work and life, place a high value into building a work environment that allows people to do awesome work, and can’t help but think about how to ensure that the right things are done right. In addition to that, I’m personable and get along well with people. I think these skills are a combination rare enough combination to help me succeed. But more importantly, such a role would integrate a lot of things I truly enjoy doing and that I’m good at. Lastly, I can treat programming as a pure hobby – it might really not be useful at all quite soon – but I truly love doing it and can treat it like playing LEGO. I could set up nvim, solidly master Python and build stuff with it, produce beautiful plots, improve my regex skills, and more for no purpose other than pure enjoyment and relaxation.

Going forward, I wanna do the following: