Stack Overflow rolled out the results of their annual 2018 survey. Over 100,000 developers participated in the survey making it a comprehensive view of software development at this moment in time, across the world. The results of this survey are broken out amongst students, professionals, and across countries. It’s worth reading the survey in its entirety.
The following are my general takeaways.
The gender and race imbalance is significant
This is a well-documented phenomenon that affects data science as it does software development generally. The problem with having these fields dominated by childless (71% childless) white (74% white) men (93% male) is that the problems being solved are the problems single white men face. Diversity, generally, is welcomed to bring new perspectives into the field. New problems and new approaches to solving those problems will strengthen the field.
There are strong advocates for diversity in the venture capital realm. Joanne Wilson (Gotham Gal) and her husband Fred (AVC ) speak and write passionately on the subject. There are many more eloquent advocates than me.
Python continues it’s parade
It’s not the most popular language but when you look at what’s beating it you’ll feel better.
Python is the most Wanted language, is off the most Dreaded list, and is third for most Loved. When thinking about what language to learn for becoming a data scientist it’s worth keeping this part of the survey in mind. Especially since other languages used for data science, R, Matlab, and Julia made their way onto the Dreaded list.
Then again as a side note, this was not a survey of data scientists so we can’t make any sweeping generalizations.
Python is a top choice on this survey for a reason.
Tensorflow and Pytorch are top of mind for developers
Individuals working close to artificial intelligence are most concerned with equity, not killer robots.
It might say more about how difficult it’s going to be to create the singularity but more data scientists are concerned about matters of equity. As we should be especially since recent news has brought bad actors in the data space.
Data science offers lucrative salaries requiring few(er) years of experience in development
I found the following chart intriguing. Data scientists have fewer years of experience but are receiving higher wages. When you stop and reflect it makes sense. Hidden in here is a warning signal to folks who run or manage data science teams. Data scientists do not typically have robust software engineering experience. A software engineer should be embedded in these teams to help architect the solutions data scientists devise.
Data scientists are responsible for their deeds
It’s not enough to say you were just doing as your employer directed. People don’t buy that excuse any more than they did after WWII. We are each responsible for the work we produce. Data scientists, by way of creating these powerful models, can have a greater impact on wider society. The ability to reach and influence people in these ways necessitates that we develop and ethics we can adhere to.
Data science is going mainstream
This is a meta takeaway. It’s interesting to me that a significant portion of this survey focused on data science and artificial intelligence. It could have focused on other areas of software engineering and development — design, augmented reality, dev-ops but didn’t. Instead, a vast chunk of precious real estate was used to explore data science.
The future looks very bright for the field.