Last week, I attended a Dataiku screening of “Data Science Pioneers,” a documentary about data science and its role in our evolving technological society. The documentary brought to light many valid points regarding society’s perspective of data science, the field itself and the ethics surrounding it which I wanted to highlight in this post.
The term “Data Science” seems so new to us, but in reality, the field has been around for a very long time. Data Science is rooted in statistics, and topics like Artificial Intelligence (AI) and Machine Learning (ML) date back to the 1940s from mathematician Alan Turing. Data science was mainly used in private research for many years, but the creation of the internet introduced more data than ever before; meaning that we had to figure out what to do with it.
With data growing across fields and industries, this brought rise to so many different applications of data science. Even with every thing that data scientists have discovered and created up until now, there is still so much unknown possibilities out there. As its applications that impact society expands, data scientists need to be able to do the math and the technical work and also be able to communicate with stakeholders as well. If we cannot translate findings to those who need to make important decisions, then what is the point of investing in data science?
Honestly, data scientists cannot do everything, and as the demand for these skills increase the term “data scientist” has begun to take on different definitions. You have your data scientists who strictly work on the technical aspects, but you also now have data scientists that work closer with business goals and have more business acumen (whatever the industry may be.) Therefore, moving forward, we may see more of these specialized data scientists within industry verticals versus the typical data scientists who are siloed off in a different department entirely.
There is also the ethics to consider with data science, from both working with data to potential outcomes of the products of data science. When solving a problem with data science, you look for causality to make predictions, but if your data is biased, this can lead to biased results and outcomes. Data scientists are meant to provide solutions that impact the greater society, so biased results can be dangerous. This is why humans are still important, because humans can detect bias and change their way of thinking, while machines and algorithms cannot.
One big area where this is especially important is in healthcare. You can have data on several patients who have received a treatment, but if some of these patients have confounding factors that affected their outcomes, this would introduce a bias to the data. It could be dangerous to use this data as evidence for another patient to be treated in the same way. This is why data needs to be interrogated by humans, to remove bias; something that an algorithm just cannot detect and account for.
This brings up one of the classic fears of AI, as portrayed in many movies, of robots becoming so advanced that they replace the human race. AI as we know it, is nowhere close to this happening. Machines cannot replace humans, because they cannot think in the way that a human thinks. A machine learns based on past experience, so a machine doesn’t know that you can fall off a cliff, until it actually falls off a cliff. Humans, on the other hand, know this based on common sense and intuition. So no need to worry, because we are not going anywhere. AI and ML should be seen as something that enhances us rather than something that replaces us.
While concepts like AI and ML may seem like something that is out of this world, in reality it affects our everyday lives. From recommended products on Amazon, to chatbots on websites and improved processes in airports, AI and ML is already everywhere, and there is still so much possible. We cannot run out of work to do or things to improve, because at the end of the day, Data Science is here to optimize and further improve the quality of everyday life.
If this post peaked your interest in Data Science, head on over to Data Science Pioneers to find a screening near you.