SA
메인 내용으로 이동

Can we ever build TikTok for Text

A month ago, Kevin Systrom and Mike Krieger, the co-founders of Instagram, launched a new app called Artifact. Artifact is a personalized news feed app that uses machine learning to understand users' interests and offer a feed of popular articles from a curated list of publishers. The app is described as a TikTok for text, where users tap on articles that interest them, and Artifact will serve similar posts and stories in the future. There have been other attempts to create personalized news feeds, including Google's Home feed, Facebook's Suggested for You, and Apple's News app. Even I also tried once. However, none has been as successful as TikTok in creating an engaging and addictive user experience.

While I most admire Kevin Systrom and Mike Krieger, it's a flop this time. They attempted to solve the problem of needing more good reads in recent days. They sought a solution by recommending good texts, but I see this as an inherent problem of reading texts. Reading and parsing articles take much more mental energy, making it difficult for text-based apps to be as engaging as video-based apps like TikTok.

There have been numerous attempts to resolve this, such as:

  • Chat interface. Chat-like ebook services like Hooked use text message format, presenting stories in short, bite-sized messages.
  • Summary Bot. Include tl;dr bots to summarize news articles automatically.
  • Interactive Articles. Use interactive components like touchable graphs and scroll-reacting elements to make text-based content more engaging and interactive.
  • Illustrations. Hire an illustrator or use generative AI to create visual elements for text-based content.
  • ... Or Give Up. Run full-time short-form YouTube or TikTok channels rather than trying to make texts more engaging.

Artifact, as of right now, is fundamentally not different from any other news app. All internet companies now have some algorithmic recommendations, and I felt Artifact is no better. They should've done something more breaking, for example, a NewsGPT.

Perhaps Systrom and Krieger are bibliophiles who wanted to reintroduce long-form texts' beauty. I agree; I am also a bibliophile. But we must understand that the notion of texts changes over time. For example, when Confucius was alive, texts were cryptic and poetic, and writing and distributing a commentary was largely forbidden as it undermined critical thinking. They will gasp at modern-day books' detail and kindness if they read the current books.

Similarly, some might also see a modern, practically illiterate generation who cannot finish a single book in one sitting as the undermining of human intelligence. While I also lament the shortening attention span of humans, this is a current of history that we cannot alter by pure will. Instead, we need to learn how to coexist with such changes in human media and preserve intelligence, as seen in 효과적 가속주의.