Connect with us

Technology

The 20 Internet Giants That Rule the Web

Published

on

The 20 Internet Giants That Rule the Web

Can I share this graphic?
Yes. Visualizations are free to share and post in their original form across the web—even for publishers. Please link back to this page and attribute Visual Capitalist.
When do I need a license?
Licenses are required for some commercial uses, translations, or layout modifications. You can even whitelabel our visualizations. Explore your options.
Interested in this piece?
Click here to license this visualization.

The 20 Internet Giants That Rule the Web (1998-Today)

With each passing year, an increasingly large segment of the population no longer remembers images loading a single pixel row at a time, the earsplitting sound of a 56k modem, or the early domination of web portals.

Many of the top websites in 1998 were news aggregators or search portals, which are easy concepts to understand. Today, brand touch-points are often spread out between devices (e.g. mobile apps vs. desktop) and a myriad of services and sub-brands (e.g. Facebook’s constellation of apps). As a result, the world’s biggest websites are complex, interconnected web properties.

The visualization above, which primarily uses data from ComScore’s U.S. Multi-Platform Properties ranking, looks at which of the internet giants have evolved to stay on top, and which have faded into internet lore.

America Moves Online

For millions of curious people the late ’90s, the iconic AOL compact disc was the key that opened the door to the World Wide Web. At its peak, an estimated 35 million people accessed the internet using AOL, and the company rode the Dotcom bubble to dizzying heights, reaching a valuation of $222 billion dollars in 1999.

AOL’s brand may not carry the caché it once did, but the brand never completely faded into obscurity. The company continually evolved, finally merging with Yahoo after Verizon acquired both of the legendary online brands. Verizon had high hopes for the company—called Oath—to evolve into a “third option” for advertisers and users who were fed up with Google and Facebook.

Sadly, those ambitions did not materialize as planned. In 2019, Oath was renamed Verizon Media, and was eventually sold once again in 2021.

A City of Gifs and Web Logs

As internet usage began to reach critical mass, web hosts such as AngelFire and GeoCities made it easy for people to create a new home on the Web.

GeoCities, in particular, made a huge impact on the early internet, hosting millions of websites and giving people a way to actually participate in creating online content. If it were a physical community of “home” pages, it would’ve been the third largest city in America, after Los Angeles.

This early online community was at risk of being erased permanently when GeoCities was finally shuttered by Yahoo in 2009, but luckily, the nonprofit Internet Archive took special efforts to create a thorough record of GeoCities-hosted pages.

From A to Z

In December of 1998, long before Amazon became the well-oiled retail machine we know today, the company was in the midst of a massive holiday season crunch.

In the real world, employees were pulling long hours and even sleeping in cars to keep the goods flowing, while online, Amazon.com had become one of the biggest sites on the internet as people began to get comfortable with the idea of purchasing goods online. Demand surged as the company began to expand their offering beyond books.

Amazon.com has grown to be the most successful merchant on the Internet.

– New York Times (1998)

Digital Magazine Rack

Meredith will be an unfamiliar brand to many people looking at today’s top 20 list. While Meredith may not be a household name, the company controlled many of the country’s most popular magazine brands (People, AllRecipes, Martha Stewart, Health, etc.) including their sizable digital footprints. The company also owned a slew of local television networks around the United States.

After its acquisition of Time Inc. in 2017, Meredith became the largest magazine publisher in the world. Since then, however, Meredith has divested many of its most valuable assets (Time, Sports Illustrated, Fortune). In December 2021, Meredith merged with IAC’s Dotdash.

“Hey, Google”

When people have burning questions, they increasingly turn to the internet for answers, but the diversity of sources for those answers is shrinking.

Even as recently as 2013, we can see that About.com, Ask.com, and Answers.com were still among the biggest websites in America. Today though, Google appears to have cemented its status as a universal wellspring of answers.

As smart speakers and voice assistants continue penetrate the market and influence search behavior, Google is unlikely to face any near-term competition from any company not already in the top 20 list.

New Kids on the Block

Social media has long since outgrown its fad stage and is now a common digital thread connecting people across the world. While Facebook rapidly jumped into the top 20 by 2007, other social media infused brands took longer to grow into internet giants.

By 2018, Twitter, Snapchat, and Facebook’s umbrella of platforms were all in the top 20, and you can see a more detailed and up-to-date breakdown of the social media universe here.

A Tangled Web

Today’s internet giants have evolved far beyond their ancestors from two decades ago. Many of the companies in the top 20 run numerous platforms and content streams, and more often than not, they are not household names.

A few, such as Mediavine and CafeMedia, are services that manage ads. Others manage content distribution, such as music, or manage a constellation of smaller media properties, as is the case with Hearst.

Lastly, there are still the tech giants. Remarkably, three of the top five web properties were in the top 20 list in 1998. In the fast-paced digital ecosystem, that’s some remarkable staying power.

This article was inspired by an earlier work by Philip Bump, published in the Washington Post.

Click for Comments

AI

Charted: The Exponential Growth in AI Computation

In eight decades, artificial intelligence has moved from purview of science fiction to reality. Here’s a quick history of AI computation.

Published

on

A cropped version of the time series chart showing the creation of machine learning systems on the x-axis and the amount of AI computation they used on the y-axis measured in FLOPs.

Charted: The Exponential Growth in AI Computation

Electronic computers had barely been around for a decade in the 1940s, before experiments with AI began. Now we have AI models that can write poetry and generate images from textual prompts. But what’s led to such exponential growth in such a short time?

This chart from Our World in Data tracks the history of AI through the amount of computation power used to train an AI model, using data from Epoch AI.

The Three Eras of AI Computation

In the 1950s, American mathematician Claude Shannon trained a robotic mouse called Theseus to navigate a maze and remember its course—the first apparent artificial learning of any kind.

Theseus was built on 40 floating point operations (FLOPs), a unit of measurement used to count the number of basic arithmetic operations (addition, subtraction, multiplication, or division) that a computer or processor can perform in one second.

ℹ️ FLOPs are often used as a metric to measure the computational performance of computer hardware. The higher the FLOP count, the higher computation, the more powerful the system.

Computation power, availability of training data, and algorithms are the three main ingredients to AI progress. And for the first few decades of AI advances, compute, which is the computational power needed to train an AI model, grew according to Moore’s Law.

PeriodEraCompute Doubling
1950–2010Pre-Deep Learning18–24 months
2010–2016Deep Learning5–7 months
2016–2022Large-scale models11 months

Source: “Compute Trends Across Three Eras of Machine Learning” by Sevilla et. al, 2022.

However, at the start of the Deep Learning Era, heralded by AlexNet (an image recognition AI) in 2012, that doubling timeframe shortened considerably to six months, as researchers invested more in computation and processors.

With the emergence of AlphaGo in 2015—a computer program that beat a human professional Go player—researchers have identified a third era: that of the large-scale AI models whose computation needs dwarf all previous AI systems.

Predicting AI Computation Progress

Looking back at the only the last decade itself, compute has grown so tremendously it’s difficult to comprehend.

For example, the compute used to train Minerva, an AI which can solve complex math problems, is nearly 6 million times that which was used to train AlexNet 10 years ago.

Here’s a list of important AI models through history and the amount of compute used to train them.

AIYearFLOPs
Theseus195040
Perceptron Mark I1957–58695,000
Neocognitron1980228 million
NetTalk198781 billion
TD-Gammon199218 trillion
NPLM20031.1 petaFLOPs
AlexNet2012470 petaFLOPs
AlphaGo20161.9 million petaFLOPs
GPT-32020314 million petaFLOPs
Minerva20222.7 billion petaFLOPs

Note: One petaFLOP = one quadrillion FLOPs. Source: “Compute Trends Across Three Eras of Machine Learning” by Sevilla et. al, 2022.

The result of this growth in computation, along with the availability of massive data sets and better algorithms, has yielded a lot of AI progress in seemingly very little time. Now AI doesn’t just match, but also beats human performance in many areas.

It’s difficult to say if the same pace of computation growth will be maintained. Large-scale models require increasingly more compute power to train, and if computation doesn’t continue to ramp up it could slow down progress. Exhausting all the data currently available for training AI models could also impede the development and implementation of new models.

However with all the funding poured into AI recently, perhaps more breakthroughs are around the corner—like matching the computation power of the human brain.

Where Does This Data Come From?

Source: “Compute Trends Across Three Eras of Machine Learning” by Sevilla et. al, 2022.

Note: The time estimated to for computation to double can vary depending on different research attempts, including Amodei and Hernandez (2018) and Lyzhov (2021). This article is based on our source’s findings. Please see their full paper for further details. Furthermore, the authors are cognizant of the framing concerns with deeming an AI model “regular-sized” or “large-sized” and said further research is needed in the area.

Methodology: The authors of the paper used two methods to determine the amount of compute used to train AI Models: counting the number of operations and tracking GPU time. Both approaches have drawbacks, namely: a lack of transparency with training processes and severe complexity as ML models grow.

Continue Reading

Subscribe

Popular