I have been using Follow intermittently for a while.
Initially, I came across an article on a public account titled "Zhang Yiming's Disdain for RSS, Now It's Hard to Get a Code," which instantly piqued my interest. I couldn't wait to buy an invitation code and subscribed to a few channels on Follow.
At first, it felt quite novel; the interactive interface was very simple, and the functionality was decent.
I became interested in the product design itself, so I browsed the blogs of DIYGod and the founder of RSS3. I read a lot about their views and hopes regarding decentralization, web3, and blockchain, as well as the Follow product.
Follow hopes to break the information cocoon through information aggregation and anti-algorithms, creating the next generation of browsers.
Follow has two highlights: one is based on RSShub, and the other is mutual subscriptions, allowing you to see what others are looking at. According to the author of Rss3, "capitalizing information distribution."
Zhang Yiming's evaluation of RSS back then was that even with information aggregation, there was simply too much information to consume. Lei Jun's evaluation of RSS was that it wasn't worth spending money on this for such a small group; it had no prospects.
The information cocoon is not the original intention of algorithmic push. Algorithmic push primarily addresses the issue of information overload through the filter of interests. I don't have a strong need for RSS, nor do I enjoy scrolling through Douyin or Xiaohongshu. Opening Follow every few days shows a pile of information, which is indeed a bit overwhelming. I believe that information overload is the main contradiction, while the information cocoon is a secondary contradiction.
Currently, information distribution is basically divided into two types: algorithmic distribution and social distribution. If algorithmic distribution provides users with information within their information circle, then social distribution serves as the driving force for users to step outside their information circle. The most efficient way to obtain information is actually to pay attention to the information that the people you follow are viewing. Follow has made a breakthrough in this regard; in the past, such social distribution usually existed within strong acquaintance social networks, like WeChat. Follow brings this distribution into unfamiliar social contexts while also emphasizing user privacy.
I don't believe that we must oppose algorithms; algorithms can serve as the first layer of filtering. The relationship chains and networks within social contexts can act as distributors after the filtering. Interpersonal relationships are often hierarchical, centered around oneself, with information flowing outward like ripples in water. Each user acts as a distribution center, and the places where such ripples resonate may be where information needs to go.
Moreover, the logic of "capitalizing information distribution" actually lies between user-generated content and user-shared content, where users reorganize and distribute content. This can indeed be a form of distribution, somewhat like an AI public account that creates information gaps? In contrast, the freedom and threshold for such distribution have been significantly lowered, making content distribution more refined and interesting.
We must acknowledge that information has indeed become a form of capital. Mastering one's information flow is akin to holding capital.
As for the topic of decentralization, I won't comment too much. History has proven that one strict party falls, and another strict party rises; the process of decentralization will inevitably produce new centers. However, history progresses through such repetitions and changes.
The above is merely my personal opinion; if there are any inaccuracies, please point them out. I welcome discussions.