Daniel Lemire's blog

, 4 min read

Being shallow is rational

Pundits often lament who people have become shallow. They no longer sit down to read books cover from cover. Instead of writing thoughtful 2-page emails, they write a single line. Sometimes they do not even write it themselves, as they often delegate to an artificial intelligence like Google’s Inbox.

In The myth of the unavoidable specialization, I argued that far from heading toward a world where everyone is a narrow specialist, we are headed toward a world of hypergeneralists. And what hypergeneralists do is to surf on the surface as fast as they possibly can. Hypergeneralists do not spend 3 weeks reading one book. They skim 3 books a day.

I believe that hypergeneralists are onto something. Our world is characterized by three attributes:

It is fast changing. Entire new fields are created and destroyed every few years. The economy turns around every decade. When you spend three months studying deeply one topic, you have to consider that it could have greatly diminished value in a few years.

At this point, people typically suggest that you can invest in topics that do not change. However, it is a lot harder than you might think to predict what won’t change. In the 1990s, there was no sign that newspapers would become obsolete. In fact, newspaper owners had every reason to think that their golden years were ahead of them. Only ten years ago, programmers would have considered that investing in Microsoft expertise was a safe bet. Those who did so missed the whole mobile revolution and they mostly missed the cloud revolution as well. They still can write beautiful desktop apps, but nobody cares, not even Microsoft.- It is vast. For the last few centuries, it has been impossible for any human being to read everything that is being written. But it has now gotten to the point where no matter who narrowly you define a domain, you cannot possibly hope to keep up even if you do nothing all day but read your peers.

It is varied and interconnected. Maybe you think that studying ancient Greece or Group theory will make you immune to obsolescence. But people who study ancient Greece using virtual reality these days to walk the streets of Athens. And I bet that there are people applying the emerging new field of deep learning to Group theory.

You should always be ready to learn a new programming language, a new concept borrowed from sociology and a new statistical test. Your mind needs to remain agile.

Computers can always provide you with the details. If I need, this morning, to learn everything there is to learn about a specific form of cancer or about a new programming language, it is easy. It is easy as long as my mind is prepared for it.

What is a prepared mind?

  • You need to be agile. You should be able to go from thinking like a physicist to thinking like a programmer within a few hours. Sometimes you need to cover many different roles quickly, sometimes you need to go deep into one specific role. This requires you to have received various training.
  • Your memory is an index, not an encyclopedia. What is important is not how quickly you can remember facts out of thin air, but how quickly you can look things up. To look things up, you need to know that they even exist. Your mind must be aware of many things, but it does not need to store the details.
  • You must be able to process information quickly using constantly tuned filters. Our brains are not great at thinking deeply and quickly. It is one or the other. But you can use a good set of heuristics to guide you. In effect, you must develop good filters. Your filters need to be constantly adjusted, as you risk being blind to important new facts… but you cannot live effectively without good filters.- You need to constantly expand your mind with people and tools. No matter who you are, your naked brain is not smart enough. Trying to make it alone, without great tools, is trying to get around on foot. You can be the greatest athlete on the planet, you still benefit from having access to a car and to planes. If you are not connected to super smart people, you also cannot win. The smart crowd knows more than you do.

And this sums up the hypergeneralist: agile thinking, memory like an index, finely tuned filters and mind expansion using people and tools.

Let us forget about the old man living in a monastery, reading thick books in isolation. He is the scholar of the past, not the bright mind of the future.