Perennial Narrative: New and Shiny vs Tried and True
How Robert Shiller's Narrative Economics also drives the narratives in tech.
tl;dr
Narrative Economics asserts that narratives drive the economic market
Certain narratives are perennial and revolve around two opposing stories.
A perennial tech narrative is new and shiny versus tried and true.
Narrative Economics
Narratives are a fundamental part of the human experience. We love to tell others stories, and we indulge in crafting stories for ourselves. Nobel Laureate Robert Shiller’s new book Narrative Economics: How Stories Go Viral and Drive Major Economic Events explores the link between narratives and the economy over the past two centuries.
Being relevant to the times, Shiller starts the book by exploring Bitcoin and its explosive growth. He claims that narratives fueled this explosion. Ironically Shiller wrote, “Practically no one, outside of computer science departments, can explain how cryptocurrencies work.” And yet, everyone seems to want a piece of Bitcoin. The fad was not fueled by the revolutionary technology rather by the narrative, like the tulip mania of the mid-1600s
Narratives spread like an epidemic
Professor Shiller saw a connection between narratives and epidemics. Stories spread and die off like a disease propagating through a closed population. To capture how fast or slow and big or small the population gets infected, Shiller referred to the Kermack-McKendrick SIR Model, an epidemiological model invented in the 1920s. The model tracks populations of susceptible individuals, infected individuals, and recovered individuals over time. We can construct an infection curve as shown above by modifying the disease’s contagion rate and recovery rate. The claim is that narratives too follow this curve.
Google Books Ngram
Unlike a disease where you can retrieve hospitalization logs, Professor Shiller needed a way to capture how popular a narrative was to the people at the time. Books and texts were the perfect proxy. As a story gets popular, people write about them.
For over 15 years, Google has scanned through millions of books in multiple languages since the 1800s. With this corpus of data, Google Books Ngrams gives us the power to look up the occurrences of a word relative to time. While we cannot search for a story directly, keywords around a story are fair game. Using this tool, Shiller could quantitatively gather data and produce curves like the above. Then, he could compare similarities between stories like the spikes of interest in bimetallism versus bitcoin.
The big insight was that new stories were often not truly unique; instead, it is a twist of an older narrative with different actors in a different setting revolving around two competing themes. Professor Shiller coined this a perennial economic narrative. In the book, he describes nine more perennial narratives that occurred over the past century.
New and Shiny Versus Tried and True—a Perennial Tech Narrative
Inspired by the book, I wondered if an underlying perennial narrative also drove the technology space.
I started by looking for stories. The first that came to mind was a group of developers talking over a new programming language in a heated discussion.
“The Rust is the best programming language, and everyone should use Rust!” — an ex-colleague
As powerful as Google Ngrams was, it could not handle words with multiple meanings. For example, Google Ngrams’s search cannot differentiate between Rust the language, Rust the game, or rust on a piece of iron.
I switched to Google Trends to tackle this limitation, which can break down a common keyword into subcategories. While not directly querying through published works, search queries seem good enough at capturing interest.
Unlike Ngrams, which returns a frequency percentage, Google Trends returns a weighted index between the given keywords. In the examples below, I chose to replot the data without the weights in cases where I want to focus on the curve patterns.
We can see a grow-peak-decay phase for the different languages over the past three decades by combining the two graphs. The Go language curve, with its peak and decline, is the most identifiable to me. Around 2014, I would see blog posts or posts that ported an existing project to Go daily. Fast forward half a decade, while the language is still great and widely adopted, the hype and number of such posts have dropped. Today, I feel Rust is following the same cycle.
Chasing the next shiny thing
Early in my career, I focused on front-end development. I started with jQuery, then quickly picked up Backbone.js. When Angular.js came out, I jumped ship. Next, when React was all the hype, I moved to work solely in React. Unconsciously, I chased the next new shiny thing and boasted with little anecdotes about how much my productivity increased.
The veterans around me were more skeptical, and they took their time before adopting the new tech. Instead, they waited and stayed with the tried and true technology they knew by heart until the new tech was “ready.”
A little wiser today, I have shifted from the new and shiny camp to the tried and true. This perennial narrative captures how these technologies evolve and grow for me and the industry as a whole.
Even companies seem to follow the same trends. The chart above clearly shows that people forgot about Myspace when they moved to Facebook (Meta). When Instagram was breaking out, Meta made a well-timed call to acquire the company. Looking at it today, the acquisition in 2012 perfectly complimented Meta’s grip on social media dominance. If you were bored with Facebook, here’s a new and shiny app, Instagram. Given TikTok’s recent meteoric rise, it’s not surprising that Meta is worried.
Buzzwords and tech fads, too, follow the curve. While the technology underlying the trend is real, the behavior to throw this at all problems does end. Some might be happy to hear that I think we are past the peak big data hype. We might also be approaching peak deep learning and machine learning hype.
Conclusion
A perennial tech narrative is new and shiny versus tried and true. We all know that person would try out all the latest tech, and we also all know someone who is still using decades-old technology. Even you can switch sides at different points in your career, and this is natural. This eternal conflict is a necessity to drive innovation and produce progress.
Taking the learnings of the SIR model, we can say with certainty that technology’s hype/popularity has a beginning, a peak, and an end. While it is not always clear something has peaked, it is pretty obvious once it is in decline. We make predictions about where a particular technology is heading.
I hope this mental model can help more people understand one another and guide the decision-making process of whether to pursue new and shiny technology.
Interesting insights while researching…
You can have projects be really infectious yet very forgettable. Google has managed to produce a number of these. The world talks about them for a week, then they are forgotten.
Apple has mastered the game of keeping its products relevant over time. The yearly WWDC events inject a new spike year after year.
Reddit just keeps growing as if there’s no cure from Reddit!