As newer fields emerge within data science and the research is still hard to grasp, sometimes it’s best to talk to the experts and pioneers of the field. Recently, we spoke with Adam Ross Nelson, data science career coach and author of “How to Become a Data Scientist” and “Confident Data Science.” In the interview, we talked about what confident data science is, how data scientists can confidently and ethically use AI, and emerging fields like prompt engineering. You can listen to the full Lightning Interview here, and read the transcript for two interesting questions with Adam Ross Nelson below.
Q: What is Confident Data Science?
Adam Ross Nelson: You know, I’ve put a lot of thought into that, not only because of the book itself, but when I was approached last summer to do this book as part of a series, I thought to myself “What exactly is confident data science?” One of the things to know that helps make sense of the title is that this book is part of a series you can find other books in the series. There’s a confident UX book, a confident coding book, and then my book is on confident data science.
I put some thought into that, and I think confident data science connects really well with the 80-20 rule. That means that roughly eighty percent of the work is done with 20 percent of the tools. So if in data science, where we have many common tasks, challenges, and use cases, and data science presents a full range of solutions, all of which – or eighty percent of which – can be serviced by 20 percent of the tools, and 20 percent of the knowledge associated with data science. Then, data scientists have to specialize a little bit above and beyond that 20 percent, but no single data scientist could ever know a hundred percent of the theme and I think that’s true for many fields.
For me, confident data science is acknowledging this 80-20 rule, acknowledging this 80-20 dynamic. Just knowing that data science is a very diverse field, and that there’s room for many paths with many backgrounds and with many interests, and no one data scientist could ever know the entire field. It’s knowing what you know and doing it well.
Q: What problems arose that led you to write this book?
Adam Ross Nelson: I saw a problem in the data science publishing world a couple of years ago, when I went to my bookshelf and I pulled roughly 10 books from my bookshelf – all on data science – and some of them were very broad data science and some of them were very specific to an industry. Then, I combed through the appendices the indices and the glossaries and I looked for ethics. And would you believe only two books of the ten mentioned or discussed ethics in any meaningful substantive, deep way?
So right there I knew that there was an opportunity, and I knew then if I ever wrote a book that I would I would work to correct that and bring a book together that was more intentional about centering ethics, responsibility, and trustworthiness, throughout and infused the book.
The other thing I did for this book was rewrite the history of data science. A lot of folks will point back to the mid-1990s as some of the origins of data science and some of the first folks who really invented data science, but I point to Ada Lovelace and Florence Nightingale in the 1800s. If you’ve never read about Ada Lovelace, know that she was essentially writing about generative AI in the 1800s. She wrote about how she was working with Charles Babbage on the computing engine, which was an early computer in many views. Ada Lovelace is often credited with being the first computer scientist because she is often credited with being the first to write a computer algorithm. In her journal, she writes about how the analytical engine might be able to compose music, and that’s generative AI right there.
How to learn more about confident data science and responsible AI
While many data scientists may have a good heart and only the best intentions when developing algorithms, sometimes it’s hard to ensure that they’re using AI confidently, ethically, and responsibly. By attending ODSC West this October 30th to November 2nd, and specifically checking out the Responsible AI track, you’ll learn everything you need to know about ethically implementing AI.