Summary
For much of the past two decades, the internet has moved toward concentration. A relatively small number of platforms became the main routes through which people search, discover, watch, shop, communicate, and encounter culture. Google organized access to information, YouTube became a dominant home for online video, Amazon became a starting point for shopping, and social platforms increasingly determined what people encountered without requiring them to decide where to go first.
These systems created enormous convenience, but they also changed the character of the internet. Billions of people now move through many of the same interfaces, recommendation systems, content formats, and commercial environments. Publishers, creators, and businesses adapt to the incentives of those systems, while generative AI makes it possible to produce competent content, imagery, software, and experiences at a scale that was previously difficult or expensive.
Abundance may eventually create its own counterpressure. People have always used the internet for more than efficiency. They gather around interests, develop communities, follow people whose taste they trust, participate in shared activities, discover unfamiliar things, and return to places that feel relevant to who they are. As the largest parts of the internet become more automated and increasingly capable of anticipating what people want, smaller and more distinctive destinations may become valuable precisely because they offer something different.
AI could make those destinations easier to create. The same technology that can summarize the web can also reduce the cost of building software, producing media, personalizing experiences, and serving relatively specialized audiences. The result could be an internet that becomes more concentrated at the level of discovery while becoming more fragmented at the level of experience.
That would create a larger possible role for the website. Instead of functioning primarily as a collection of pages, a publishing surface, or a place to transact, it could become an owned digital product where people learn, participate, use tools, meet others, buy things, and connect their digital interests to the physical world.
NewBrains proposes six elements for thinking about these destinations: Content, Community, Conversation, Commerce, Capability, and Connection. The objective is not for every website to contain all six. It is to ask what combination could turn a website from somewhere people are sent into somewhere they have a reason to return.
The internet became a system of highways
The early web was fragmented by design. People moved between individual websites, forums, blogs, directories, stores, publishers, and communities, often finding one place through another. The experience could be inefficient and difficult to navigate, but much of the identity of the internet came from the individual destinations within it.
Search engines and social platforms made that internet considerably easier to navigate. Google created a powerful way to organize an expanding web, while Facebook and later Instagram, TikTok, and other social platforms centralized the discovery of people, media, and culture. YouTube became a dominant destination for online video, while Amazon brought an enormous range of products into a single commercial environment.
Over time, these platforms became something like the highway system of the internet. They made it possible to move through an enormous digital world without knowing much about the individual places inside it. A creator could reach an audience without building distribution from scratch, a small retailer could reach customers through search or social advertising, and someone interested in an obscure subject could find information or other people who shared that interest within seconds.
The convenience came with a structural change. As more activity moved through a relatively small number of platforms, those platforms increasingly influenced what succeeded within them. Publishers learned how to appear in search results, creators learned which formats travelled through recommendation systems, retailers adapted to marketplaces, and advertisers developed creative around the conventions of individual platforms. None of this eliminated the enormous diversity of the internet, but more of that diversity began travelling through the same roads.
An abundant internet can also become a more familiar one
The largest digital platforms face an unusual problem: they have vastly more things to show people than anyone could possibly consume. Recommendation systems help solve that problem by deciding which small portion of the available supply deserves attention.
Those systems can also create powerful incentives for the people producing what gets recommended. A successful format can be reproduced thousands of times, while a particular style of headline, thumbnail, video, product presentation, or creative execution can spread quickly once people understand that it performs. The result is not an internet in which everything becomes identical, but one in which many different subjects and cultures are increasingly expressed through familiar formats.
Generative AI could deepen that tension because it changes the economics of production itself. Writing, imagery, video, software, research, advertising, personalization, and interfaces can increasingly be produced with less time and fewer resources. That expansion will create things that could not previously have existed, but early research also suggests that widespread use of the same models can reduce diversity across large collections of creative work. One 2025 study comparing human and GPT-generated writing across 2,200 college admissions essays found that additional human-written essays contributed substantially more new ideas to the collective pool than additional GPT-generated essays, even when researchers modified prompts and model parameters to encourage greater diversity. [1]
The implication is not that AI inevitably makes culture less creative. Models can produce unusual work, people can use them in very different ways, and new tools frequently create new forms of expression alongside the conventions that eventually develop around them. The more relevant possibility is that when competent production becomes abundant, competence itself becomes less differentiating. An internet containing dramatically more content and software may make distinctiveness, judgment, culture, and recognizable points of view more valuable rather than less.
People still look for places that feel different
There is a human reason to believe that greater abundance will not simply produce an appetite for more efficient recommendation.
People have used the internet to gather around interests almost from the beginning. In 2001, Pew Research Center found that 84 percent of American internet users had contacted or obtained information from an online group, while 49 percent of those participants said the internet had helped them connect with people or groups that shared their interests. These communities formed around professions, hobbies, sports, beliefs, lifestyles, local organizations, and other areas of shared interest. Pew described some of them as virtual versions of the “third places” people inhabit outside home and work. [2]
The technology surrounding that behaviour has changed repeatedly, from forums and message boards to Facebook Groups, Reddit communities, Discord servers, creator audiences, multiplayer worlds, group chats, and countless other forms. What has remained more consistent is the underlying desire to find other people who care about the same things.
That desire is not only social. People also browse because they are curious, follow ideas without knowing exactly where they will lead, develop expertise by spending time inside a subject, and build tastes that cannot always be reduced to an initial query or recommendation. The internet is useful when it removes friction from something we already know we want, but some of its value has always come from giving us places where we can spend time without knowing precisely what we are going to find.
Large platforms can support these behaviours, and many of the internet’s most vibrant communities already exist inside them. The question is whether the increasing scale, automation, and abundance of those environments could also strengthen the appeal of smaller places with a clearer sense of who and what they are for.
People may continue using the highways while becoming more interested in the destinations.
AI could make a more fragmented internet possible
At first, AI appears likely to push the internet toward even greater concentration. If an assistant can search across websites, interpret what it finds, compare alternatives, summarize information, and eventually complete more tasks, people may have fewer reasons to navigate individual websites themselves.
Early behaviour around AI search already points toward that possibility. Pew Research Center analyzed the browsing activity of 900 U.S. adults in March 2025 and found that users clicked a traditional Google search result on 8 percent of visits when an AI summary appeared, compared with 15 percent when no summary was present. Links contained within the AI summaries were clicked on only 1 percent of visits, while users were also more likely to end their browsing session after encountering an AI-generated summary. [3]
This creates a plausible future in which a handful of AI interfaces become even more important gateways to information and services. The websites underneath those systems would still matter as sources of information, products, transactions, and authority, but people might interact with them less frequently for routine informational tasks.
The same technology is producing another change that could pull in the opposite direction. AI is reducing the resources required to create many of the digital experiences that sit underneath those gateways. Software provides one measurable example. A 2026 quasi-experimental study involving developers at Ant Group found that adoption of an AI coding system increased code output by more than 50 percent. [4] Similar changes are occurring across design, media production, research, personalization, customer interaction, and other activities required to build and operate digital products.
If it becomes substantially easier to create and maintain sophisticated digital experiences, scale becomes a different calculation. A useful destination may not need hundreds of millions of people if it can serve two million serious runners, 100,000 architects, 50,000 watch collectors, or a particular community of gardeners, musicians, investors, gamers, or parents.
The large platforms would not need to disappear for these places to matter. In fact, they could remain the primary way people discover them. The more interesting possibility is an internet that becomes more concentrated at the top and more fragmented underneath: increasingly powerful AI systems and platforms that help people navigate the world, alongside many smaller destinations designed around particular interests, communities, and behaviours.
The return of the destination
For years, much of digital strategy has treated the website as somewhere people are sent. A search result leads to an article, an advertisement leads to a landing page, a social post leads to a product, and an email leads to an offer. The website sits at the end of distribution systems that are largely controlled elsewhere, and its job is often to complete whatever intention caused someone to click.
That model remains useful, but it represents only one possible role for an owned digital experience. If the internet becomes more fragmented around interests and communities, some websites could begin functioning less like endpoints and more like destinations that people know, choose to enter, and have reasons to revisit.
The distinction matters because a destination serves a different purpose from a feed. A feed continually brings things to us, while a destination is somewhere we decide to go. The reason for making that choice can come from expertise, entertainment, community, identity, taste, useful tools, access to people, or simply the expectation that spending time there will produce something worthwhile.
This is already how many strong communities work. A serious runner does not return to a running community because it contains one particular article about improving a 10K time. The value comes from the larger relationship with running: learning about it, talking about it, tracking progress, meeting other runners, participating in events, discovering equipment, and developing as a runner over time.
If websites begin playing a larger role in those relationships, the distinction between a website and a digital product starts to become less useful.
The website can become something bigger
Most websites still reflect the organizational categories from which they were created. Publishers publish content, retailers sell products, B2B companies explain services and collect leads, while community, customer service, events, applications, and other experiences frequently sit elsewhere.
An interest-based destination does not necessarily need to preserve those boundaries because people do not experience their interests in such clean categories. Someone who cares deeply about running may read about training, compare shoes, track a run, follow athletes, talk with other runners, join a local club, enter a race, and eventually purchase something. From the runner’s perspective, these are not separate digital disciplines. They are different parts of being a runner.
The opportunity is therefore larger than adding more features to a conventional website. It is to think of the website as an owned digital product that can bring together different parts of an audience’s relationship with an interest. Information still matters and commerce still matters, but neither necessarily needs to define the entire experience.
This way of thinking also places an important constraint on the idea. A company cannot manufacture an interest simply because it wants people to spend more time with the brand. The strongest opportunities exist where a business already has a credible relationship with something people care about independently of the business itself.
People do not run because Nike exists. Nike matters because people run.
The same principle applies elsewhere. People do not play guitar because Fender needs engagement, build things because LEGO needs traffic, or care about beauty because Sephora needs conversion. These businesses participate in interests and behaviours that already have their own cultures, communities, rituals, knowledge, and reasons for returning.
That is the foundation on which a destination can be built.
The Six Cs of an owned digital destination
If a website is understood as a product built around an audience’s recurring interest, the practical question becomes what that product might contain. NewBrains proposes six overlapping forms of value that can help turn an owned digital experience into a destination: Content, Community, Conversation, Commerce, Capability, and Connection.
Content gives people something worth learning, watching, reading, or discovering. It can include education, entertainment, expertise, inspiration, research, stories, and perspectives, but its role is larger than filling a publishing calendar. In a destination, content helps people go deeper into the interest that brought them there.
Community creates a relationship between the people who share that interest. This might involve groups, members, creators, customers, experts, enthusiasts, or other forms of participation. The important distinction is that the audience can become aware of and derive value from one another rather than maintaining separate one-to-one relationships with the brand.
Conversation allows the destination to respond. People can exchange ideas with one another, ask experts questions, interact with creators, communicate with the business, or increasingly use AI to explore information and receive guidance. As generative interfaces become more capable, conversation may become one of the ways static websites begin behaving more like responsive products.
Commerce allows products and services to exist naturally inside the interest rather than requiring the entire experience to organize itself around a transaction. A product can become relevant because of what someone is learning, making, training for, planning, or trying to accomplish.
Capability gives people something useful to do. Calculators, training plans, trackers, personalization, configurators, creation tools, planning systems, saved progress, diagnostics, and other utilities can make the destination valuable even when someone has no immediate intention to consume content or purchase something. This is also where the distinction between a website and a product becomes particularly visible.
Connection allows the digital relationship to extend beyond the screen. Events, classes, races, stores, meetups, clubs, consultations, workshops, and other physical experiences can connect an online interest to the places and people through which it is lived in the real world.
These six elements should not become six mandatory sections on every website. Some businesses may have compelling reasons to use only two or three, while others may have opportunities across most of them. Their usefulness lies in changing the question from what pages a website needs to what kinds of value could give a particular audience a reason to spend time there and return.
What the Six Cs could look like in practice
Nike provides a useful hypothetical example because the company already operates across many of these areas. Imagine Nike.com not primarily as Nike’s online store, but as one of the main digital destinations for people who run.
Training guidance, athlete stories, and running culture could provide Content, while runners, clubs, and shared challenges could create Community. Coaches, experts, other runners, and AI-supported guidance could contribute to Conversation. Shoes, apparel, race equipment, and other products would remain the Commerce layer, but would sit within a larger relationship with running. Training plans, run tracking, progress, personalization, and planning could provide Capability, while local running clubs, races, stores, and events could create Connection between the digital destination and running in the physical world.
The important change is not that Nike would have six different types of functionality. It is that the experience could be organized around running rather than around Nike’s organizational structure or the immediate objective of selling a shoe. Someone might return because they are training for a race and encounter commerce naturally as part of that relationship.
Fender could apply the same model around playing guitar without producing anything that resembles the Nike experience. Lessons, musicians, songs, and technique could sit alongside communities of players, access to instructors and experts, practice tools, sound experimentation, events, and instruments. The recurring behaviour would not be visiting Fender. It would be playing music, with Fender finding useful ways to participate in that behaviour.
LEGO provides another variation because building naturally combines consumption and creation. A destination could help people discover ideas and techniques, share what they make, participate in challenges, meet other builders, use digital creation tools, attend events, and buy sets or individual pieces when they become relevant to a project. The commercial relationship would remain important, but it would exist inside a larger culture of making.
The model becomes less convincing when the underlying interest is weak. Not every insurance company needs a community, every utility needs an editorial platform, or every household product needs to become a lifestyle. The Six Cs are useful only after a business has identified a genuine behaviour, interest, problem, profession, or culture in which it has permission to participate.
A destination needs a reason to return
This requirement may ultimately matter more than any individual element of the framework. Adding community software does not create a community, publishing articles does not create a publication, and adding an AI assistant does not automatically turn a website into a useful product.
The strongest destinations attach themselves to behaviours that already recur in people’s lives. A runner runs, a musician plays, a builder makes, a home cook cooks, and an investor researches, monitors, and makes decisions. Each return creates an opportunity for the digital product to become more useful because the person’s relationship with the underlying interest is continuing.
This is why the idea differs from the familiar argument that every brand should become a media company. Publishing more material may increase reach, but reach alone does not create a destination. A destination becomes meaningful when the business can contribute something useful to the activity people came to pursue, whether that contribution is knowledge, people, tools, products, access, experiences, or some combination of them.
The question for a business therefore becomes less about how to create engagement with the brand and more about what the brand can contribute to an interest people already care about. That shift may sound subtle, but it changes the organizing principle of the website. The brand no longer needs to sit at the centre of every interaction. The interest can.
What this means for digital leaders
The immediate response to AI has understandably focused on discoverability. Businesses are asking how their information appears in AI systems, whether LLMs understand their products, how generative search affects traffic, and what needs to change as more customers encounter information without entering the website from which it originated.
Those questions matter because AI is likely to become an important intermediary between businesses and their audiences. Product information, policies, expertise, research, inventory, services, and other forms of structured knowledge will increasingly need to work outside the boundaries of the website itself.
The other side of the problem receives less attention. If people need to visit websites less often for routine information, the moments when they deliberately choose to visit an owned experience become more important. Digital leaders therefore need to think not only about how information leaves the website, but about what remains uniquely valuable when someone enters it.
The Six Cs provide one way to examine that opportunity. A business can look at the behaviours already taking place around its category and ask where people currently learn, gather, talk, buy, use tools, and meet in the physical world. Much of that activity may continue to happen on YouTube, Reddit, TikTok, Discord, specialist applications, marketplaces, creator channels, physical stores, and other environments. The objective should not be to drag every interaction back into an owned platform.
The more useful question is whether some of those behaviours become more valuable when they are connected. A company with genuine expertise, useful tools, an active community, distinctive products, and physical experiences may discover that treating each one as a separate channel leaves considerable value on the table.
That requires a different ambition for the website. Traffic and conversion will remain important, but they can sit alongside measures such as direct visits, repeat use, contribution, participation, membership, tool adoption, and the strength of the relationship people develop with the destination over time.
The website stops being only the endpoint of somebody else’s distribution system and starts becoming a product worth seeking out in its own right.
The internet beyond the highway
The largest technology platforms are unlikely to disappear, and AI may make some of them even more important. People will continue wanting convenient ways to search, discover, watch, communicate, shop, and receive answers, while AI assistants may become another major layer through which the wider internet is navigated.
But a highway becomes more useful when there are places worth travelling to.
The same technological shift that is making aggregation more powerful is also making creation easier. The same abundance that allows an AI system to answer an extraordinary range of questions could increase the value of environments with recognizable taste, expertise, relationships, culture, and community. Greater convenience at one layer of the internet does not require greater sameness at every layer underneath it.
The history of online communities also suggests that the desire to gather around shared interests is not an artifact of any particular platform. People have repeatedly rebuilt that behaviour as technology has changed, moving through forums, blogs, social networks, group chats, multiplayer environments, creator communities, and other forms. The interfaces evolve, but people continue looking for others who care about the same things and for places where those interests can be explored more deeply.
AI may therefore produce a more complicated internet than either the optimistic or pessimistic versions of its future suggest. We could depend on a smaller number of increasingly intelligent systems to navigate the digital world while simultaneously spending time in a larger number of specialized places once we get there.
For businesses with a legitimate role in those interests, that creates a different way to think about the website. It does not have to be only somewhere people read about the company, find information, or make a purchase. It can become the place where content, community, conversation, commerce, capability, and connection come together around something people already care about.
The internet is getting smarter. The opportunity for the website may be to become something bigger.
The Idea
NewBrains proposes two related ideas from this analysis.
The fragmented destination internet describes a possible internet in which discovery becomes increasingly concentrated through large platforms and AI systems while experiences become more fragmented across specialized destinations organized around particular interests, communities, and behaviours.
The website as product describes the role an owned digital experience can play within that environment. Rather than treating the website primarily as a collection of pages, publishing surface, or storefront, the model treats it as a digital product built around a recurring audience interest or behaviour.
NewBrains’ Six Cs of an owned digital destination provide a practical framework for considering what that product might contain:
Content — something worth learning, watching, reading, or discovering.
Community — relationships between people who share the interest.
Conversation — interaction between audiences, experts, creators, brands, and AI.
Commerce — products and services situated naturally within the interest.
Capability — tools and functionality that help people do something useful.
Connection — ways the digital experience extends into people, places, and experiences in the physical world.
Not every destination requires all six. The framework is intended to help businesses identify the combination that can create meaningful value around an interest their audience already has.
If AI reduces the need to visit websites simply to retrieve information, the strategic question for the website changes. Being findable will still matter, but it may no longer be enough.
The larger question is whether there is something worth coming for and a reason to return.
Sources
- Homogenizing Effect of Large Language Models (LLMs) on Creative Diversity: An Empirical Comparison of Human and ChatGPT Writing. Computers in Human Behavior: Artificial Humans, 2025. Three preregistered studies analyzing 2,200 college admissions essays and comparing collective semantic diversity in human and GPT-generated writing.
- Pew Research Center. Online Communities. 2001. Research examining participation in online communities, shared-interest groups, local communities, and the internet as a virtual “third place.”
- Pew Research Center. Google Users Are Less Likely to Click on Links When an AI Summary Appears in the Results. July 2025. Analysis of browsing behaviour from 900 U.S. adults examining searches with and without AI-generated summaries.
- Gambacorta, Leonardo, Han Qiu, Shuo Shan and Daniel M. Rees. Generative AI and Labour Productivity: A Quasi Experiment on Coding. Journal of Financial Stability, 2026. Quasi-experimental research examining the effect of generative AI adoption on software-development productivity.
