Are Algorithms Killing Innovation? NewsWire

NewsWire: 8/8/22

• Are algorithms quashing creativity and risk-taking? In an op-ed, columnist Ross Douthat argues that algorithm-driven operating models are driving entire industries towards homogeneity. (The New York Times)
• NH: In just the past three years, fast-fashion juggernaut Shein has become the third-most valuable startup in the world. In part, its breakneck growth is credited to powerful algorithms that are able to collect vast amounts of user data, analyze what’s selling, and help producers churn out an average of 6,000 (yes, 6,000) new designs every day. When shopping on Shein, the sheer number of search results can be overwhelming.
• Yet something strange is happening. In theory, algorithms like Shein’s allow for hypertargeted personalization, with users choosing from literally millions of items and receiving recommendations catered to their tastes. Yet at a time when producers have the capacity to customize every aspect of their work and Americans have more consumer choices than ever, the products and services for sale look more more and more similar.
• In his editorial, Douthat points out that in the music industry, the hottest area of investment is not up-and-coming artists, but the back catalogues of Boomer and Silent musicians. (See “Streaming Has Transformed the Music Charts.”) Today's pop hits are becoming less musically diverse, with a growing share of songs attributed to a shrinking pool of writers and producers. Meanwhile, Hollywood has been overtaken by reboots and sequels. (See “Whatever Happened to the Middlebrow Movie?”)
• In an age of unlimited personalization, we’re seeing more and more homogeneity. Why?
• Two forces are pushing in the same direction. The first is happening on the consumer side. Today, consumers are being served up algorithmic recommendations that are based on what other people with similar demographic traits are doing. Netflix auto-plays shows it thinks we will like. Social media sites organize their feeds in a way that keeps us scrolling as long as possible. Our favorite stores, restaurants, and supermarkets send us deals and suggest products according to our buying habits. Google Maps sorts its driving routes according to variables like traffic patterns and gas usage. We are constantly guided towards what we’re being told to want, regardless of whether we want it or not.
• And on the flip side, creators, business owners, and marketers are increasingly trying to “write to the algorithm.” The average person, of course, has no idea how these algorithms work. But businesses know that something will help them appear earlier in search results, show up in ads, and be recommended more often, and they’re constantly trying to guess what it is in order to succeed. Often, this involves running their own tests--basically, writing algorithms to figure out the algorithms. This experience has given rise to the term “algorithmic anxiety”: business owners feeling like their success is being controlled by some unseen, enigmatic digital overlord.
• As a result, businesses--just like consumers--are being incentivized to move to the middle. Algorithms simply double down on what’s working and produce more of the same. This doesn’t mean that nothing new ever gets created, but the system is not set up to encourage this to happen. Although Shein releases a mind-boggling number of clothes, they are mostly slight variations on previous styles: a puff sleeve here, a different pattern there. Although there are millions of YouTubers and TikTokers, the language and emotional register they use are uncannily similar. (“Hit that like button and subscribe!”)
• This logic applies to more than just products and brands. There’s a direct analogy, for instance, between algorithmic thinking and the rising popularity of huge ETFs in investing, in which largely passive investors follow the herd. (See “Will Index Funds Kill Competitive Capitalism?”) They are always going to do as well as the whole market, but avoid the bigger risk of going out on their own.
• Of course, people can opt for safe options even if they aren’t using technology. That’s risk aversion. What algorithms do is make avoiding risk effortless. They are good at maximizing old ideas within known parameters, but they can’t create better ideas or anticipate a future that is unlike the past.
• Generational forces have helped further entrench algorithmic thinking. Millennials are drawn to community and are comfortable following the herd. Xers are drawn to hard numbers and data-driven insights. Boomers, by contrast, have always insisted that following their own creative instincts is superior to an algorithm--and many will tell you that this is why cultural artifacts from the 1960s and '70s have withstood the test of time. But as a generation, Boomers are losing power. 
• Indeed, many of the cultural touchstones that are resurfacing today are from previous generations--back when a hit movie was just a hit movie and a bestselling book was just a bestselling book. Last year, 9 of out of the 10 top-grossing movies were sequels or part of a franchise. When something original does come along, it's milked for all it's worth...making it even less likely that will be room for the next original invention to break through. Every new idea spawns an endless tide of spinoffs and branded tie-in products. Why come up with something new when it’s safer to produce an echo?