My thoughts on "Science in the age of selfies"

Hongtao Hao / 2020-10-17


I encountered this opinion article, Science in the age of selfies , published on PNAS when I searched for “selfies” on Google Scholar. I thought it was about selfies, but it was not at all.

Although not about a topic I thought it was, this opinion is very well written and has some insightful ideas, at least for me. I will in the following talk about what I thought about several major arguments in this article and what I learned from them.

Major Ideas and My Thoughts #

1. The speed of scientific discoveries is slowing down #

The authors believed that the last fifty years’ scientific discoveries are not as exciting as those made between 1915 and 1965. I am not at a good position to evaluate this argument since I am not very familiar with the natural sciences community.

2. Too much communication and too easy collaborations #

I agree. In my mind, mingling with other scholars at conferences is good to keep updated on the scientific trends. However, as the two authors warned, too much communication might end up people doing the same thing. It is true that nowadays, artificial intelligence and machine learning are prevailingly popular – so popular that they might be stifling new ideas.

This reminds me of an American scientist1 who says that he won’t do the most popular science. The fact that it is popular by no means indicates its importance. He suggests doing research that you consider significant, whether it’s popular or not.

In terms of collaboration, I also don’t believe that all great ideas emerge from groups or teams. Yihui Xie, who came with Rmarkdown and Bookdoww, tools that changed the way how people use R and write documents, also mentioned in one of his blog posts2 that he did the work by working alone.

For me, teams are great for implementing ideas, not for initiating ideas. For example, Fei-Fei Li and her colleagues completed the work of ImageNet by virtual of around 50,000 people from 167 countries on Amazon Turk. This, of course, is an extreme case, but it illustrates that teams are useful when there exists a good idea, which, unfortunately, do not start from teams most of the time.

3. Pressure to publish a lot might inhibit deep thinking #

This I also agree with but am not sure how likely this will change. From a practical standpoint, if we don’t count the number of publications and their citations, how can we evaluate the performance of a researcher objectively? If, as the two authors suggested, performance is judged by quality rather than quantity, who will be the judge? And how can we make sure that their evaluations are both accurate and fair?

What I can learn from this article: #

  1. Try my best to maintain prolonged concentration. Stay away from distractions. Deep work is not possible without deep thinking.

  2. Be careful of popular research. Find out what I really care, and are fascinated about, and work hard on it, rather than popular things that appear to be able to make me successful, which might be an illusion.

  3. Keep a balance between quantity and quality in terms of publication. I want to make contributions to the scientific community but at the same time I don’t want my future wife and kids to starve.


  1. Sorry that I cannot remember his name. ↩︎

  2. Sorry that I cannot remember which post. ↩︎

Last modified on 2020-10-22