Hello, dear readers! I am releasing an early blog this week as I will be out of town over the weekend celebrating a family wedding. I will be back next week with another entry (provided fall courses don’t eat up all of my free time!). If you haven’t already, check out my two previous pieces here and here.

This blog was inspired by “Seattle Data Guy” Benjamin Rogojan’s insightful work which you can find here. Also these two incredibly probing and poignant articles in Fortune and The New Yorker.

First, I would like to say that I am a huge fan of machine learning being applied to healthcare/biology. I am such a fan, in fact, that I married it… Well, sort of. My husband is getting his PhD in Bioinformatics and Integrative Genomics and has a particular fascination for deep learning. He’s the smartest person I know, and you should follow him on Twitter @amarotaylorw.

As much as I want to see the promise of machine learning come to fruition, Rogojan makes some strong points about the issues surrounding its application in healthcare.

Machine learning algorithms and models are becoming somewhat of a household name in many industries. Given the hype surrounding AI, there is a level of trust in the technology at this point that far exceeds our understanding of it or its capabilities. This is precisely the issue that concerns Rogojan. He posits this blind trust in machine learning could transfer into its applications in healthcare, leading providers to spend less and less time questioning the output. The prospect is chilling, isn’t it? Physicians as automatons, feeding the insatiable appetites of these algorithms with more and more data.

It is easy for us non-data scientists to become concerned about this future to a degree that stops us from moving forward. But is it as scary as we might be inclined to believe?As Rogojan’s article poses, relying on machine learning for diagnostic purposes could result in doctors losing touch with some of their current core competencies. He compares this to people using GPS to get around. How many of us can read a map now? How many of us rely on our own eyes and memory of a city alone to find our way? I would guess not many.

“I do often wonder if we as humans have the discipline to not rely on [machines] for the final say.”

It is a matter of responsibility on the behalf of the user, as much as the data scientist behind the algorithm. The tools they create will shape how healthcare providers interact with the system and the decisions they ultimately make.

I do not believe people like my husband are out there trying to made physicians obsolete. They are trying to answer the medical questions that plague mankind in new, better ways. They aim to enhance the capabilities of modern medicine and most of them recognize that humans in general, and doctors in particular, bring some irreplaceable ingredients to the mix.

A human touch is crucial in something as delicate and critical as healthcare. Medical problems are complex, they are not just a collection of symptoms. Diseases and treatments vary from patient to patient, and the way a person describes their illness to a doctor can change drastically depending on a variety of factors. As Siddhartha Mukherjee points out in this amazing piece for The New Yorker, the level of engagement of a physician during care can have important therapeutic value all on its own.

“I noticed other patterns in Bordone’s interactions with her patients. For one thing, they almost always left feeling better. They had been touched and scrutinized; a conversation took place. Even the naming of lesions—“nevus,” “keratosis”—was an emollient: there was something deeply reassuring about the process. The woman who’d had the skin exam left looking fresh and unburdened, her anxiety exfoliated.”

I believe these things can happily coexist and enhance each other but, for that to happen, the people behind the machines have to ask themselves tough ethical and moral questions about the impact of their work. These problems haven’t become evident as of yet due to the fact that the technology is still developing, and these systems are not yet as easy to use as asking Siri to give you directions to the nearest pharmacy. But we will get there eventually, and we should not wait for that day to think about these questions. I however, have hope that most data scientists and engineers are carefully pondering these issues today.

The use of machine learning doesn’t necessarily mean the obliteration of the medical profession. This technology has the potential to empower physicians, to allow them the time, space and resources to ask more probing questions. It can enable them to apply their creativity to the many currently unsolved problems in healthcare. Provided the leaders of today take the time to look at the future through a human lens.