This is a reflection on the past and present of the Internet.
<h1 toutiao-origin="h2" >01 Bit mirror image of the Internet and the world</h1>
I've had a very simple question with many friends, in your eyes, what is the Internet?
To answer this question, you must first answer another question, why did the Internet appear?
In my opinion, the world is made up of three elements: matter, energy, and information. So what is information?
According to the definition given by CE Shannon, the grandfather of informatics, information is an uncertain description of the state of motion and the way things exist.
The process of human access to information is the process of eliminating uncertainty. Because the concept of describing information confusion and uncertainty is "information entropy", you can also think of the process of human acquisition of information as a process of "information entropy reduction".
The entropy of binary sources of information
In other words, access to information can make you go from "not understanding" to "understanding".
In order to survive in this world, our genes have "designed" us to be an efficient information acquisition and processing system. For example, we (or all primates) have evolved the ability to recognize colors, and our eyeballs can recognize the spectrum in the 380 to 780 nm band, which gives us the ability to distinguish food by color, find cover, and perceive danger.
Of course, this is only a very basic ability to obtain information, in the process of human evolution for millions of years, in order to meet the needs of our social organization, we have evolved a more complex and advanced ability to obtain and understand abstract information, which is the ability to process information such as words, pictures, music and so on.
Simply put, access to information is important to us, but what does this have to do with the Internet?
Here we also need to understand two concepts, one is the signal and the other is the channel.
Transmission of Information
The publisher of the information (the source) publishes a message that is carried in some physical medium (this is the signal), transmitted through the medium (this is the channel), and received by the recipient of the information (the letter).
For example, a chestnut is that you say a sentence (this sentence contains information), this sentence is transformed into a sound wave signal, which is transmitted to me through the channel medium of air, and I hear it, which is a process of information transmission. In this process, there will be "noise" in the channel, the noise will interfere with the signal, the information transmission will be biased, and finally even lead to the failure of information transmission. Therefore, different signal bearing forms and different channels will have relative physical limits of information transmission. For example, if you say a word 1 km away, I may not hear it.
Therefore, the distance that the signal can be transmitted without interference has a great impact on the efficiency of how much social organization we can maintain.
So in order to maximize the preservation of signals from interference, as far as possible to transmit / obtain information over the longest distance, we humans have invented various technologies. To date, the most advanced technological tool for conveying information is the "Internet."
If you have studied informatics in college, basically the first lesson will be about this.
But if we only understand the level of tools, our understanding of the Internet is not deep enough. Because when the Internet transmits information, it is not a one-way transmission, in fact, the Internet retains the ability of the real world for everyone to publish and obtain information, so the Internet forms a "network". The network extracts as much as possible all the figurative objective and abstract subjective information known in the world, maps it to the Internet, and then transmits it in electrical signals (the speed of light).
So, I often say that the Internet is a bit (unit of information) mirror image of the real world.
And the information on this bit image is transmitted at the "speed of light".
S=VT
We all know this simple formula, which means distance = speed * time. Simply put, if information is transmitted at the speed of light, then the radius distance from which we obtain the information is very long in the same amount of time. How long has it grown? You've definitely heard of the concept of a "global village," and in the decade when the Internet first came to China, people especially liked to say it. Although it is a bit earthy, this sentence is indeed very graphic. The earth became a village.
It was good in the beginning, our radius for information expanded, and we completed an "evolution" with computers and the Internet. But we may not be able to get used to this change, because the environment facing human beings has changed from "lack of information" to "information overload" for the first time in the 10,000 years since the birth of civilization.
In fact, in the development process of the Internet, we can sort out two main veins, namely:
More and more information on the Network;
More and more people are surfing the Internet.
The two main veins themselves promote each other to form a closed loop, that is, more and more people are making, publishing, and involving more and more information on the Internet, and more and more information on the Internet also attracts more and more people to use it.
We all know that by 2019, the number of Internet users in the world will be exactly 4 billion, of which less than 900 million are Chinese Netizens. So do you know how much information there is on the Internet?
Source: IDC, 2017 Data Age 2025 white paper
In 2017, IDC (International Data Corporation) did a study, they estimated that there were about 20 ZB of information on the Internet at that time (1 ZB = 1 trillion GB), and predicted that this data would probably double to 40 ZB by 2019.
Global data volume for 2019
So a rough calculation, each netizen is theoretically distributed to an average of 1ZB of information. Of course, the reality is not egalitarianism, it is only easy to understand here. Suffice it to say that even if we exercise this egalitarianism, you will never be able to fully accept all this information in your short life. What's more, in the ocean of information that can be reused, the reality is that what you need to find in the 40tzB of information sea for the rest of your life today is only the dozens of terabytes that you need.
The information you can get exceeds the limit of information you can handle, which is information overload.
The pioneers of the Internet actually realized this problem for a long time.
The world's first website
http://info.cern.ch/hypertext/WWW/TheProject.html
Tim Berners-Lee, the father of the World Wide Web (WWW), invented the World Wide Web at the same time as the world's first website, which used hypertext technology to connect CERN (European Nuclear Research Center) laboratories, because of the use of hypertext (that is, later links), so people can easily browse the aggregated information. Tim Berners-Lee unveiled the project on August 6, 1991, a day that is also considered the birth of the World Wide Web. This website introduces the specifications of hypertext, the details of website construction, the installation and use of browsers, etc., and later this website also includes some other website examples, so it is also considered to be the world's first website directory.
It was also from this day that everyone connected to the World Wide Web (or, in other words, the Internet you are familiar with) has the right to build their own graphical website and corresponding http website. Http URLs, which then use more natural spelling, are already much easier to access than previous FTP addresses.
Of course, in the era of the World Wide Web, although Tim Berners-Lee had tried to do some directory work on World Wide Web sites, http sites had such a huge explosion since then that the job obviously needed a dedicated agency to serve. So from that era onwards, the long road of Internet information distribution was opened.
We can roughly divide the mode of Information Distribution on the Internet into four eras according to the popular period of its dominant mode of information distribution:
Categorical Index - Portal Era;
Search Engines - Search Times;
Subscribe to The Watch - SNS Era;
Recommendation Algorithm - Feed Era;
In addition to these four eras, there is also a long-standing "high-heat update - community hot post" model (this model has also become a very important information distribution model in China because of the rise of Baidu Tieba in the search era). It is worth noting that most of the information distribution models born in these eras are produced with the corresponding technological changes and applications, and behind them are accompanied by the complete upgrading and innovation of business models and the rise of a network era.
The evolution of these information distribution models has shaped the Internet giants of different eras. However, these models are not completely replaced, more is the new information distribution model backwards compatible with the earlier model, and then through the innovation of the business model, the latter-led enterprises press the predecessors to the ground friction (at least to some extent, to break the leadership status of the predecessors).
<h1 toutiao-origin="h2" >02 Category Index - Portal Era & Search Engines - Search Times</h1>
Although the birth of BBS forum and search engine technology is earlier than the birth of the World Wide Web, but the application of productivity needs to adapt to the objective laws of the development of the times, the birth of the World Wide Web, the dominant information distribution in the network world is not the search engine technology and BBS forum, but the classification index based on hypertext technology.
Although the first Internet classified indexing service was not Yahoo, the most well-known and successful case is indeed it.
The first version of Yahoo in 1994!
Founded by Chinese entrepreneur Jerry Yang, the site quickly defeated his main rival in the mid-1990s with efficient management and aggressive marketing strategies to become the world's most important online portal (yes, the so-called four major portals in China - Sina, Sohu, NetEase, and Tencent were originally Yahoo!). copycat). The concept of the so-called "portal" may be a little unfamiliar to today's young people, but for people at that time, this word is very vividly depicted, which is the first step for most people to go online.
It may be hard for people today to imagine how people could start if they were looking for "information" if they were looking for something "information" 25 years ago, when there were no search engines. So Yahoo! This manual inductive classification, manual entry and collection of classification search directory website, is where you can quickly find out what new websites appear on the Internet. You can think of these sites as "yellow pages" (number books with phone numbers for businesses), or in a way, Yahoo! It is the network yellow pages (Ma Yun first wanted to do the network yellow pages, and later took Yang Jeryuan's investment, not without reason).
The human imagination does not come out of nothing. We will always make some guesses about the future based on some existing things, combined with some changing trends. For people who have never seen the future of today, it makes perfect sense to "materialize" a website that aggregates all the URL information, refer to the existing telephone yellow pages, and "skeuomorphize" the design product.
And Yahoo! The business model can be said to be very rough and simple, that is, to sell advertising, or more precisely, to sell banner advertising on the website.
Yahoo in 1997! Car banner ads that appear above the search box
Yahoo at the time! To increase revenue, it is also very simple to add banner slots on the homepage (including the large category pages of the various index catalog pages that later index the catalog pages) (increase Ad Loads), and sell individual ads more expensively (increase Ad Price), for which Yahoo! Willing to give advertising sales a very high commission, while placing a large number of market advertisements, in order to incentivize sales and education advertisers.
Yahoo in 2002! A bunch of ads on the home page banner
Early Yahoo! After the success of the information distribution model of the classification index, the form of the product was rapidly evolved, and the search function was first added in 1995, but this search is not the same as the later search engine, mainly used to do the quick search of the classification search, and the search results often point directly to a certain URL (this detail is particularly important when talking about the search engine later). Then Yahoo in '96! The mailbox service was started again, and then the business tentacles were expanded to news information services (this service is equivalent to benchmarking the business of media newspapers).
As the world's largest Internet portal, web directory, email address and news information website at that time, with the rapid growth of the number of internet users around the world, Yahoo's market value once exceeded 100 billion US dollars, so that in today's Google and China's four major portals, there is always Yahoo! The sense of both vision.
Then, the search engine that would end the portal era in the future appeared.
Strictly speaking, search engine technology was born a year before the World Wide Web, and the predecessor of the modern search engine was Archie, which was born in the 90s, a technology that finds files from FTP hosts by file name. In 1994, the first search engine using spider crawling technology in the modern sense was born, Lycos (the word Lycos is the name of a tarantula). 4 years later, the Google we are familiar with was also born.
The initial difference between these search engines and portals for categorical indexes is the different ways in which web information is obtained. The classification index is manually entered, while the search engine uses a spider crawler to crawl information fully automatically. In addition, manually classified directories mainly contain URLs, while spider crawlers can crawl each specific web page. For users, the former needs to enter the website to continue to find information, and the latter can be in one step, convenient and fast.
Obviously, search engines are much more efficient at obtaining web information than categorical indexes. So why didn't people use search engines so much in the first place? Because although the early search engines could climb a lot of things, in a large number of related results, the search engine could not accurately "guess" which is the result you want.
The reason for Google's rapid rise is not only because of their tireless and fast crawling of web pages or the concise design of products. The early winners came from Google's creative redefinition of the way search results are sorted.
One of Google's two founders, Larry Page, invented the PageRank algorithm — a measure of the value of a page by the number of links to it. Google has made a difference in optimizing search results. In fact, including the PageRank algorithm and the Hilitop algorithm, HITS algorithm, TrustRank algorithm, and SandBox (sandbox) used to deal with some bad web pages, google became the company that paid the most attention to the efficiency of information distribution around 2000.
However, although the early Google was technologically advanced, as a technology company, they could not yet compete with Yahoo, which had a large number of users and a huge amount of cash! Direct head-to-head confrontation. In fact, not only did they not confront, Google even Yahoo in 2002-2004! Exclusive outsourcing of search technology. Yahoo with Google technology! The experience is rapidly improving, the number of clicks is soaring, and the giant Yahoo! Happy to lean forward and backward, completely unaware that this little brother will become his own gravedigger in the future.
As Yahoo! Google's Larry Page once wanted to sell Google's predecessor, BackRub, to Yahoo! And by 2002, Yahoo! Also once tried to buy Google for $3 billion, if it weren't for Google's counter-offer to $5 billion, Yahoo! The expensive abandoned the deal, and the history of the world's Internet was almost rewritten (such stories are really constantly playing out on the Internet).
The first version of Google in '98 imitated Yahoo! also had an exclamation point
Google's early team photos
Really let Google do it on Yahoo! Beyond that is the innovative business model AdWords (keyword advertising). Google makes every searched keyword a potential advertising vehicle, and every search results page becomes a source of advertising revenue.
As we said earlier, the dynastic change in the era of Internet information distribution is mainly around two factors:
Advances in the application of information distribution technologies;
Revolutionary upgrade of business models;
In Yahoo! The ad was originally sold according to the popularity of the banner position of the page view popularity. This business model is actually no different from selling advertisements for telephone yellow pages, or selling newspaper advertisements. The maximum revenue from selling ads mainly depends on how many pages a user can visit, and the popularity of different pages determines the price level. In order to sell more ads, you have to add an advertising area in the best position. And the homepage is the best location, which is why the homepages of all the early portals look as crowded as the city center.
Then in order to further increase advertising revenue, each advertising slot shortens the play time of the advertisement, plus the number of advertisements in the rotation, and then according to the popularity of the advertising position, let the sales speculate on the price of the golden advertising position and the time period, which is similar to selling TV advertisements.
In that era this was already able to serve Yahoo! Make quite a lot of money. But Google completely broke that business model.
AdWords' innovation comes from two sources:
New pricing model;
New advertising space design;
Adwords pricing is based on the user's search keywords, click on the search results to calculate the advertising revenue (that is, the famous CPC pricing). As a result, the effectiveness of advertising becomes reliable and monitorable, while the price paid at a time is very low, making the strategy of advertising very flexible, which gives small businesses that cannot afford to pay huge advertising fees the opportunity to run Internet advertising.
The ad slot of Adwords is no longer a random ad sticker and pop-up window, but directly integrated in the search results, which makes the conversion efficiency of the advertisement a qualitative leap and makes the advertiser very satisfied. At the same time, the advertising carrier has been changed, and the number of advertising slots is no longer linked to the number of pages of the portal, and there are many words in the world, theoretically, how many words correspond to the page of the advertising slot. How much money Google can make depends entirely on how many times users will search with Google, and how many times users search for Google depends entirely on how good the Google search experience is, so that the advertising revenue model and product experience are no longer opposites, but a completely consistent closed loop.
Yahoo 2004! Announced the severance of the partnership with Google, and the two went to war in full swing. Yahoo completely crushed by Google in the search business! And then Google launched the Gmail mailbox service against Yahoo! Mailbox service. Technology-leading Google soon turned Yahoo! Throw yourself behind, despite the wealthy Yahoo! Bought a lot of Internet companies to form a group army, but the result is that there is no tea egg. If not yahoo that year! Investing in Alibaba in its heyday, the company will even withdraw from our sights sooner.
Just when people thought that the war on the distribution of information on the Internet was coming to an end here. The new model is here again.
<h1 toutiao-origin="h2" >03 Subscription Watch - SNS Era</h1>
In 2004, Google vs Yahoo! When it was difficult to solve, suddenly in the United States, someone proposed a completely new concept - Web 2.0. Meaning the second generation of the Internet, the biggest difference between the second generation of the Internet and the first generation of the Internet is that Web 1.0 is led by network information service providers to provide information services, and in the Web 2.0 era, it will be led by users to provide products and services.
To be honest, this concept is actually quite vague and unreliable, but around 2004, a new generation of Internet products based on this concept was really born, many of which are widely known today: Myspace (September 2003), Facebook (February 2004), flickr (February 2004), Reddit (2005), Google Reader (2005), YouTube (December 2005), twitter (March 2006), tumblr (2007) and many more.
A common feature of these information service offerings is UGC (User Generated Content). Around 2004, the number of Internet users worldwide reached about 800 million. With the access of a large number of Internet users, the generation of Internet information is facing a low-level change.
The earliest Internet users were mainly one-way access to information services, but with the increase in the number of users accessing the network and the increase in network bandwidth, the Internet began to be able to provide more services.
The earliest BBS-derived community forum, everyone can establish their own personal homepage (yes, the earliest is a variant of the personal station, personal homepage), you can share favorite pictures of the network photo album service, you can share their own sand sculpture video network video hall...
The Internet began to shift closer to the fundamentalist meaning of the word — connecting people to people and then browsing through the information posted by individuals.
Because there has been a fundamental shift in the way information is distributed, so must the way information is distributed. In the Web 1.0 era, the shortest way to access the Internet to obtain information is through a portal or search engine that aggregates information, and now users have one more choice - through the list of users who subscribe/follow.
Whether it is a portal close to the media form or a search engine close to the tool form, when users use them, they are obtained through an intermediary, and direct subscription allows users to bypass the intermediary of information and obtain information directly from the source of information.
Let's start with the change in the way information is distributed, which is RSS.
In 1999, a new XML-based protocol, RSS, was developed, which originally represented the RDF (Resource Description Framework) Site Summary, also known as the RDF Site Summary , a technology that pushes the latest abstracts of websites that conform to RDF rules to users. Netscape , the company that invented the protocol , removed the RDF statement format , and the technology became Rich Site Summary , a technique that allows websites to push site summaries to users.
After that, Internet users in the United States began to gradually become popular in writing Weblogs (Weblogs) on personal websites, or replaced by a more familiar name, Blog - Blog. When Blog encountered RSS technology, a completely new subscription service was born - RSS collection reader.
Unlike previous portals and searches, RSS readers are completely personalized, and what users will read is entirely determined by the user's subscription list, which is an epoch-making design.
Google Reader, which is most familiar to domestic users
At the same time, on the other hand, on the side of the way information is published, although there were already fast blog building tools such as Wordpress at that time, for ordinary people, building a website is still a bit of a threshold thing.
So is there a product that can satisfy both having your own personal web page and building a personal website without understanding technology? So in 2002, the first personal homepage product, Friendster, was born, which was also the world's first social networking service (SNS).
Friendster soon ran into its biggest competitor, MySpace, and lost the competition with MySpace, even though MySpace was born as a pixel-level replica of Friendster. The original reason was simply because MySpace was a bug that allowed users to modify the HTML of the front-end UI, so users could customize the colors, wallpaper, and background of their personal homepage. This "feature" is loved by a large number of teenage users, so it instantly became a hit (yes, QQ space is MySpace's copycat). Then, under MySpace's continuous high-speed product and technology iterations, the weak Friendster was knocked out of the US market and focused on serving overseas markets (Friendster was once the most popular SNS in Southeast Asia).
But MySpace didn't get a chance to laugh too long, and we all know that Facebook, the world's largest SNS, was born in 2004. But the fledgling Facebook didn't have much of an advantage over Myspace. In fact, on the contrary, MySpace was more fashionable and cool, and MySpace, based in Hollywood, Los Angeles, was more like a weather vane of pop culture (in fact, from the perspective of people at the time, MySpce was more like the feeling of 2017's vibrato, and the starting process was very similar).
Many people think that MySpace vs. Facebook will eventually lose the battle and has a significant connection to the fact that their parent company, Intermix, was sold to News Corp. Murdoch in 2005. There's more or less truth to this view, but if we dig deeper, facebook's own efforts may be the more important reason.
Facebook and MySpace distribute social information differently, and Facebook launched a new feature in 2006: Newsfeed. The original Chinese translation of this function was called friend dynamics, but with the evolution of this function, there is now a more famous name - information flow (the early information flow and today's information flow are actually very different, which will be discussed later).
But it wasn't Facebook that invented this "streaming" feature, but Twitter (though Twitter itself may not have been aware of it at first). As mentioned earlier, RSS did a subscription function, allowing users to subscribe to the blog list to view the update of the blog, and Twitter, which was born in some later 2006, took this design a step further, directly bypassing the design of the "subscription list", and directly realizing the complete closed loop of publishing, subscribing to information, and reprinting information on the Twitter site. The presentation of information is limited to the length of 140 words (because it is short enough to facilitate the user to read and publish), and in chronological order (timeline), directly present the subscription information flow, what the user reads is also completely determined by the user, which is the origin of the earliest "information flow" design.
In addition, in addition to the design innovation of the information distribution model, some people began to upgrade the medium of information release because of the increase in bandwidth, including flickr for picture sharing and YouTube for video sharing.
The biggest threat these SNS and subscription follower services pose to search engines like Google comes from the fact that Google has no way of crawling the content of web pages as conveniently as before. If we think of information as a piece of water, then these SNS and image and video sharing sites are reservoirs, and search engines like Google are water stations. If every reservoir upstream cuts off the pipeline to the water station, the station loses some value.
In the early years of the commercial website itself has no traffic, the service is not high frequency, every time the service user will need to go from the search engine, but like SNS, its own closed-loop website with a large amount of UGC dynamic information, users have to log in almost every day, almost completely get rid of the external water station traffic import. For Google, selling traffic to Facebook has become almost a shot-and-roll deal. And the new Internet information is in the content of these reservoirs, if Google can not find a way to control some of the reservoirs, it is even difficult to say who has the greater right to speak between them and Facebook in the future.
This incident directly led to Google stopping the Google Reader business and turning to the development of Google+ (Google's own social platform, which looks very similar to Tencent doing Tencent Weibo or Microvision today). And YouTube (the reservoir of the top 3 in the world today) that has been acquired at a huge cost and subsidized huge losses for ten consecutive years.
However, the good news for search engines may be that in the era of subscription attention -SNS, although new players like SNS have begun to grasp the right to speak about incremental information, they are still very weak and insufficient in the innovation of business models. In terms of business model, SNS or social media still uses the advertising model, but their efficiency in distributing advertisements is completely incomparable with search engines, after all, advertising is also a kind of information. And the business model of the search engine is too advanced. At that time, SNS and Social Media had no advantage at all in business-related data except for the use of search engines for a long time.
That's what subscriptions pay attention to – these upstarts of the SNS era have only hit some of the dying portal news stations and upended the traditional non-Internet media industry a little more. The business base of the search engine is not shaken at all.
However, the advantage of using the time also left a dawn for the upstarts, until a new era came, and new killers were studied.
<h1 toutiao-origin="h2" >04 Recommended algorithm - Feed Era</h1>
Now we all know that the new big killer is: the recommendation algorithm. However, it should be unexpected by many people, this big killer is actually relative to the era of its big killing, and it is not much new technology.
The so-called recommendation algorithm is essentially the "algorithm that recommends information to the user" As mentioned before, Google has developed a sorting algorithm such as PageRank to distinguish which pages are more valuable, which is essentially a generalized "recommendation algorithm", because Google will rank the pages they consider more valuable at the top of the search results and recommend them to users.
The most popular "collaborative filtering algorithm" was actually born as early as the 90s, Amazon has long been using collaborative filtering algorithms, and the current use of "content-based collaborative filtering algorithms" is probably used by Amazon in 2003 (of course, Amazon was mainly based on "commodities" to do collaborative filtering, friends who are interested in the development of the recommendation system, recommended to read the book "Recommendation System Practice", the author Xiang Liang).
To say how the age of recommendation algorithms - information flow has risen, we have to start in 2006.
Facebook released a new feature NewsFeed, and the original NewsFeed, equipped with a "recommendation algorithm", the algorithm at that time called EdgeRank, the original algorithm was very rough just to do some simple weighting of different content types, but over time, Facebook began to optimize their algorithms, such as using more complex computational dimensions and adding machine learning, At that time, Facebook also copied the like design of the social networking site Friendfeed (the design of the like is an epoch-making design for the recommendation algorithm, and with the algorithm of the praise machine learning, there is an important reference for judging the satisfaction of the recommendation. Of course, in the end Facebook acquired Friendfeed in 2009, erasing threats in all senses.)
Newsfeed, which is supported by recommended algorithms, is becoming more and more useful, and users no longer consume information according to the simple subscription dimension. Less than two years after the feature went live, Facebook quickly surpassed Myspace from less than half of its daily active users to active user data. This is the first time that the recommendation algorithm has demonstrated its astonishing power in information distribution.
So the most advanced Internet information companies have begun to configure the weapon of recommendation algorithms. Whether it's YouTube, Twitter or even Netflix, which does long videos. The recommendation algorithm suddenly became a necessary feature of Internet products at that time. At this time, the recommendation algorithm - information flow, is still only a function attached to the product, completely based on the recommendation algorithm - information flow design around the product has not yet arrived at this time.
Speaking of recommendation algorithms and information flow, it is actually a very strange thing that the word feed is translated as flow. Although the use of "stream" to describe the feed is indeed quite graphic. However, when I looked up the materials, I found that my friend Pan Chao's public account flipped through the book once recorded one thing, that is, long before Zuckerberg founded Facebook, he and his classmate D'Angelo (Facebook CTO, Qura founder) wrote a music player that used the recommendation algorithm, and then applied for a patent for this personalized recommendation algorithm. The literal translation of the word Stream is "stream."
With recommendation algorithms, information flows revolutionize business models and even become product forms that can exist independently, creating a new era is only one step away. The inspiration for this step came from Pinterest. Born in 2010, Pinterest was a picture sharing tool, and the mobile Internet was about to explode in its entirety, although Pinterest was still a web product at the time.
The biggest difference between Pinterest and his predecessor Flickr is twofold:
Flickr's pictures are mainly uploaded by users, and then let users share, forward and comment; while Pinterest's pictures are not only self-uploaded, they will also use spider crawlers to crawl pictures on the network, and then identify and filter the pictures through tag classification (Pinterest's largest crawl source is flickr);
Pinterest's page that presents the result is an interactive design that scrolls down infinitely without turning the page, which people in China call "infinite scrolling" (waterfall stream).
Pinterest's waterfall design became standard on later picture sites
The former allowed Pinterest to have far more image reserves than Flickr, while the latter made an epoch-making design model for information flow products in the later mobile era.
In my 2018 essay "A Brief Analysis of Gamification", I elaborated on how the recommendation algorithm + infinite waterfall design activates the "Skinner reinforcement" of humans and makes people enter a "flow" state. And this design, with a little innovation, opens up a recommendation algorithm - information flow model can be used against the business model of search engines.
Heart flow that can be stimulated by the recommended algorithm + infinite waterfall flow design
Although the theoretical upper limit of the theoretical advertising space of the search engine is very high, the search engine has a fatal flaw, that is, the number of times a person uses the search engine is actually very limited. This limit is determined by "how much one can clearly know what one does not know." This sentence sounds a bit roundabout, and the unfolding is that people must know what they are looking for now, and then the information they are looking for, and they have to be unaware of it, so that they will have the opportunity to use search engines. In other words, search engine usage scenarios are like dictionaries (Google's umbrella company shell company in the secondary market is indeed called Alphabet, which means dictionaries), and the opportunity for people to consult dictionaries and encyclopedias is too insignificant compared to the time to do nothing.
The recommended algorithm-based products that use a completely proactive approach are specially designed for scenarios where users don't know what they should be looking for. That said, the solution isn't as efficient at distributing long-tail information, and is orders of magnitude better than an accurate search engine.
The earliest and most famous Internet information products using this design should be today's headlines. This may also be the first time that China's Internet products have led the United States in the advanced nature of product concepts in the entire history of Internet information distribution. The birth of this product may also be a watershed between China and the United States Internet copy to China to copy from China, and a concrete display of China's Internet comprehensive innovation.
In the era of recommended algorithms, China got rid of copycat
Here we also need to make it clear that many people will regard the recommendation algorithm as the only ability of the recommendation algorithm - information flow product. This perception is actually very problematic.
As I said earlier, the evolution of Internet information distribution, in fact, some similar to the evolution of organisms, with the growth of Internet users and the increase of the total amount of information on the network, we continue to need more efficient and advanced information distribution solutions. But this solution is not overarching, new technology is applied, and the past technology is directly eliminated. This process is superimposed, new technology is applied, combined with new business model innovation, superimposed into a more advanced information distribution solution, which will contain effective solutions invented in the past, and the whole is a more complete and complex system.
In fact, a product like Today's Toutiao is a comprehensive information distribution product that gathers categories, searches, follow subscriptions, and recommendation algorithms.
I believe that in the future, there will be newer and more efficient Internet information distribution products, which will also follow this "law of evolution", whether this product is in the 5G or 6G era, whether it is a mobile app or a VR/MR application.
And the Internet, the world of bit mirrors, will definitely drive us humans to continue to move forward.
< h1 toutiao-origin="h2" > written at the end</h1>
I actually wrote this article a few days ago to attend the vitality conference of today's headlines. In his opening keynote speech, Zhu Wenjia, the new CEO of Toutiao, used a diagram to depict the logic of today's headline business boundary expansion.
Today's headline business boundaries "one horizontal and one vertical"
At that time, I remembered that when I wrote "The Wild Hope of Douyin, the Crisis of Kuaishou" in 2017, I once drew a picture depicting the changes in the form of different content products from social to media.
This diagram basically reflects my understanding of the morphological changes of the entire relationship-based content. It just so happens that this axis is somewhat similar to the horizontal axis drawn by Zhu Wenjia, so I wanted to write a popular science article about how the vertical axis has developed with the times.
However, when I was halfway through writing, I found that this thing was getting longer and longer, and there were many details, so I decided to omit some details, hoping to explain this matter to a basic understanding, even if it is OK, I hope that everyone who reads it can still be satisfied.
As for how the future of Internet information distribution will evolve, such a relatively large topic, welcome to private message I leave a message to discuss.
Above all, thanks.
<h1 toutiao-origin="h2" > recommended extended reading index</h1>
Flipping through the book: Pan Chao, "WeChat is old, headlines are always young"
Flipping through the book: Pan Chao, "Recommendation is a sharp weapon, attention is a blunt weapon"
Flipping through the book: Pan Chao, "Brush TikTok, Play Fast"
Flipping through the book, Zhu Xingyu, "Why did MySpace fail miserably to Facebook?" 》
Survival in the Pen, Zhu Shiyu, "The Water Maker and the Water Deliverer"
<h2 toutiao-origin="h3" > #Invited Author#</h2>
Jin Yechen, WeChat public account: Yuanshou Jin Laohan (ID: chairmanJLH), classical Internet observer
This article was originally published in Everyone is a Product Manager. Reproduction without permission is prohibited.
The title image is from Unsplash, based on the CC0 protocol