The New Internet We can Expect, Not the One We Want

I think many who are aware of the massive problems with today's Internet have unrealistic expectations for potential replacements. First, they expect a single new network to completely replace the current Internet. Second, they expect to continue engaging in the same activities on this "new Internet", including watching hours of YouTube videos and Netflix movies every day. They also expect this new Internet to be faster, have little or no intrusive advertisements but more privacy, anonymity, security, and free speech. And, they believe this new Internet will be free--both free, as in free beer, and free, as in freedom. Unless some unforeseen technology suddenly appears, I think they are on a train headed for the land of disappointment.

According to some estimates, over the last twenty years, public use of alternative Internet networks, technically known as Internet overlay networks, has grown to include millions of users world wide. After three years of investigating alternative networks like Aether, Gemini, Gopher, I2P, IPFS, Secure Scuttlebutt, Lokinet, Tor, and ZeroNet and using ZeroNet on a daily basis, I have come to some realizations about the new Internet that we are likely to see over the next decade or two. I strongly believe that alternative networks will feature prominently, but not in the way many expect.

The New Internet We can Expect

In discussing the evolution of the Internet, an article in Forbes predicted, "Users would have increased choice in new open platforms, and the emergence of legitimate dApps would limit mass censorship. The Big Tech platforms would still exist, but they would be largely defanged as the Web 3.0 ecosystem would serve as a powerful check-and-balance." While I agree with some of this, I doubt enough current Internet users will make use of alternative networks to the extent required to "defang" current Big Tech platforms any time soon.

Please allow me to explain what I think we can reasonably expect during the next decade or two from a "new Internet". First, we can expect to keep the Internet we have today as it likely continues to worsen. It will probably grow even more controlled and splintered, thanks to increasing corporate and government interference. We can expect to continue shopping on Amazon and watching YouTube and Netflix on our current government-and-corporate-controlled Internet with intrusive advertisements, without real free speech, while being tracked and spied on. As one Hacker News user pointed out, "A ministry of truth is created when you have an organization that is deplatforming and/or censoring content. That is where we are right now; a small number of big tech companies are acting as a ministry of truth." Those who want something better, and have the knowledge required to choose wisely, can expect to have the option of using an Internet overlay network (or more likely several) for hosting their personal websites, participating anonymously and freely (as in freedom) in small-to-medium-sized peer-to-peer social networks, and communicating privately with friends and associates. In peer-to-peer networks, the computers of individual users talk to each other directly without the aid of central servers. These will be lower-bandwidth, lower-storage networks that allow users to have more privacy, anonymity, security, and free speech than they have now on the current Internet. But, they will not support many of the other features that most of us want, and a high percentage of current Internet users will most likely not use them over the next two decades.

Please let me explain in more detail what those who choose to use alternative networks can expect over the next decade or two. First, those who choose not to use them will naturally remain on the regular Internet (the HTTP/HTTPS web) where they will continue complaining vainly that they have no privacy, slow web page downloads, ubiquitous advertisements and tracking, and social media that bans them for the slightest rule violation. I believe those who educate themselves by trying alternative networks over a period of years will slowly come to agree with me. Given the technology that currently exists or that soon will, they will likely conclude that they can expect something like the following.

The Alternative Internet We can Expect over the Next 10-20 Years

Small-to-medium-sized social networks run by users, with perhaps up to 100,000 users each. Some of these will allow anyone to participate and express whatever is on their minds. Others will be moderated to various degrees. This will involve many social networks because just as on today's regular Internet, one social network will not appeal to everyone. A social network for people who merely want to watch cat videos and look at meme pictures will not exist on alternative networks because that already exists a thousand times over on the regular Internet.

Privacy networks for those who want to communicate one-on-one with people they know, both in real time and in non-real time.

Anonymous networks where individuals can display their personal websites without corporate or government interference. These will be freely available (as in freedom, not as in free beer) and practically unblockable.

Single-purpose and multi-purpose networks with applications that run on users computers that are not controlled by any company or government. One example would be a decentralized app for privately and anonymously buying and selling bitcoin. The new breed of locked-down cloud computers will not run these apps.

The above four categories of capabilities may be provided by a single network or, more likely, by a combination of networks, each providing at least one capability. These networks will be peer-to-peer Internet overlay networks because this is the only type of network that users can potentially run without outside control. These networks will not rely on a central agency that assigns IP addresses or domain names. They will hide the identities of users and protect them from being surveilled by corporations and governments. They will also be bandwidth limited because they will run on the computers and Internet connections of their users. This assumes that no one like Elon Musk appears to finance the creation some new Internet paradigm, like for example, a world-wide mesh network of inexpensive, long-range, user-programmable, wireless routers. I think we can safely assume this will not happen in the next twenty years for two reasons: 1) no strong financial incentives currently exist and 2) hardware is subject to manufacturing and import/export restrictions by governments.

Current Examples of the Limitations on the New Internet Imposed by Today's Technologies

The power of individuals to use the Internet the way it was designed to be used back in the early 1990's has been severely curtailed by corporations and governments. New technologies are being rapidly developed to fight back, but they have limitations, and the opposition is developing formidable countermeasures. At the moment, Internet overlay networks look like the most likely path to regaining the freedom that has been taken from us.

To oversimplify, an Internet overlay network is an application layer network that operates on top of the transport layer TCP Internet protocols. To access an overlay network, a user connects to the Internet and runs additional software that recognizes additional communications protocols. A few examples of overlay networks are email, Gopher, the world wide web, IPFS, I2P, and the Tor network. For example, the web uses the HTTP and HTTPS protocols to deliver web pages through ports 80 and 443 on web servers. Ports are just specific memory addresses in a computer. Email is an Internet overlay network that uses the POP3, SMTP, and IMAP protocols and operates through TCP port 25 on email servers. The Gopher overlay network uses the Gopher protocol and operates through TCP port 70 on Gopher servers. In this article, I will use the phrases "Internet overlay network" and "alternative network" interchangeably, even though we do not usually think of email and the web as alternative networks. In your mind, exclude email and the web from the phase. The meaning I am trying to convey with "alternative network" is an Internet overlay network that is not commonly used on today's Internet. Tor, I2P, IPFS, Gemini, and ZeroNet are good examples. I am sorry if this is confusing.

The good news is that we already have technologies and networks with each of the four categories of capabilities mentioned previously. However, I know of no single network that can perform all four functions or that can perform any of them well enough that hundreds of millions of regular Internet users are currently using them. None that I know of provide high levels of privacy, anonymity, speed, usability, and resistance to blocking, though some come close in some of these areas.

Perhaps the most widely-known alternative network is the Tor network. This network runs mostly on hundreds of publicly-accessible servers (AKA "Tor nodes") run by volunteers and is, therefore, not a peer-to-peer network. Partly thanks to the fact that Tor servers are limited in number and identifiable, China, Iran, Turkey, Ethiopia, Egypt, Belarus, Venezuela, and about a quarter of Russia's ISP's are either currently blocking the Tor network or have blocked it in the past. Tor users in these countries use "Tor bridges" to thwart the efforts of their governments, but if a user can find a Tor bridge, then one would think that a government should also eventually be able to find and block it.

Given the technology that is widely available to governments today, the first requirement for a network to be practically unblockable seems to be that it must have such a large number of servers (or nodes) that no government is willing to go to the expense of blocking them all. Thousands of servers are not enough. A pure peer-to-peer network currently seems to be the only type of network that obviously meets this requirement. Ultimately a government could be faced with the decision to block everyone--meaning shutting down the Internet completely--or blocking no one. Unfortunately, governments with large resources have additional means of blocking specific kinds of traffic from servers. This is a cat-and-mouse game, and the technology on both sides continues to evolve.

Most of the alternative networks I have explored are not purely peer-to-peer. Most claim to be unblockable or resistant to blocking, but they rely in some way on central servers. For example, although Secure Scuttlebutt was designed for each user to connect to a small number of his friends, it currently relies heavily on less than a dozen central servers called "pubs". Technically, this makes it a hybrid network--part peer-to-peer and part central-server-based. Social media like Diaspora and Mastodon run on servers owned by volunteers, not by peers. ZeroNet is a hybrid network that operate partly as a peer-to-peer network but also depends on Tor nodes for anonymity and on BitTorent trackers for connecting the computers of peers (nodes). I should mention that ZeroNet gives users who wish to do so the ability to set up their own trackers to increase ZeroNet's peer-to-peer character. ZeroNet nodes can also find each other on local networks without the aid of trackers. Freenet was designed to run in "darknet" mode, where users connect to other users without the use of central servers; however, it currently operates mostly in "opennet" mode, which is not peer-to-peer. It has operated this way since its inception about twenty years ago.

The major hurdle for some alternative networks to reach a pure peer-to-peer mode of operation appears to be that they have no mechanism for users to find each other. Perhaps I am missing something, but it seems to me that this should not be difficult. I assume the Distributed Hash Table (DHT) that I recently discovered is used by Tox for connecting users automatically is little more than an encrypted list of users' network addresses. A simple web page that allows users to add their network addresses to a downloadable list should enable them to assemble a globe-spanning, anonymous, peer-to-peer network themselves. I do not understand why no one has done this for networks like Freenet, Secure Scuttlebutt, and Retroshare. Similar lists of Tor exit nodes exist on the regular Internet. Are users afraid that governments could use these lists to block them? Perhaps these networks rely on software with inherent limitations that make this unworkable?

Current alternative networks also have additional technical issues which prevent hundreds of millions of regular Internet users from flocking to them. For example, although I2P is peer-to-peer and claims to be anonymous, the software is currently hard to use, and it relies on a tunneling protocol that can take hours to locate a large enough number of peers to give a user access to a fairly complete network. This makes it very awkward to use, especially without a 24-hour-a-day-connected computer. Many alternative networks also lack good search engines for locating interesting content. I do not know if technical reasons exist for this or if this is simply due to the fact that no users are willing to pay the cost of hosting search engines on their own computers.

IPFS uses a DHT to locate documents on its network. This allows it to be a pure peer-to-peer network, which makes it practically unblockable by IP address. Unfortunately, this does not make IPFS block-proof. Three methods of blocking alternative networks are based on IP address, port number, and handshake protocol. Other methods also exist. Unfortunately, IPFS is very slow, and my experience has been that most of the content is simply not accessible at any given time. The reason for this is that less popular IPFS content may only reside on a few peers in the network. If those peers happen to be off line or your computer does not have access to them, then you do not have access to the content you want. IPFS is also in the process of being taken over by Cloudflare, which is now running a large percentage of the servers on the network and is trying to take on the role of a DNS provider, under the guise of being an IPFS gateway, which I suspect hampers, if not destroys, IPFS's usefulness as a peer-to-peer network.

Although I have not yet found an alternative network that has all of the characteristics that knowledgeable Internet users should be seeking, I believe the technologies currently exist. The Tox protocol uses a DHT to connect nearby peers, and it runs fast enough to provide a grainy, pixellated video calling capability. This means the Tox protocol creates a pure peer-to-peer network that is practically unblockable by governments that are unable or unwilling to block the IP addresses of a very large number of intermittent users. The qTox client is also relatively easy to use. Unfortunately, Tox does not hide the IP addresses of users, and it only serves one of the functions listed above, end-to-end encrypted real-time communication. And, in my opinion, it does not do this very well.

I have now concluded my discussion of some of the technological limitations of current alternative networks that prevent us from having the new Internet we desire. Please allow me to discuss the limitations of the users that result in the same.

The Limitations on the New Internet Imposed by Users

I feel that regular Internet users themselves are the main impediment to the wide-spread adoption of alternative networks. Just as most of the world was slow to discover the Internet in the 1990's, they are slow to discover alternative networks today. This is partly due to widespread belief in propaganda spread by corporations and governments and partly due to users' unwillingness to understand and accept responsibility for their own digital fates. Johann Wolfgang von Goethe said, "None are more hopelessly enslaved than those who falsely believe they are free." I believe the dawning of this understanding is beginning among regular Internet users, but the learning curve from that point on is steep. Determined commitment is required to climb it, and I do not believe that level of commitment currently exists among most Internet users.

The first issue that Internet users must overcome is years of bad press about alternative networks in the mainstream news media that causes most potential users to fear and shun them. This is especially true of the Tor network. The vast majority of mainstream news articles involving the Tor network have been about criminals, illegal drugs, and child pornography. It has been portrayed as a secret network used to support illegal and immoral activities, and perhaps partly for this reason it has been dubbed a "dark network" or "darknet". While I believe this attitude is slowly changing, a recent extended email conversation with one journalist has led me to believe that this view is still a deeply held bias by many in the mainstream news media. Contrary to Tor's dark image, a 2020 study found that, "...upwards of 93% of Tor users globally go to websites on the Clear Web that are not administered anonymously and so comparatively less likely to be hosting malicious or illegal content." In other words, the vast majority of Tor users employ the Tor network for completely legitimate reasons of privacy and anonymity.

Another problem that must be overcome is unrealistic expectations. Most users of the regular Internet react in a number of predictable ways when they first encounter an alternative network. Usually, they make an annoyed face and say, "Where are my YouTube videos and shopping opportunities? This place is a cesspool of Nazis, Jew haters, and Misogynists. This network sucks, and I'm leaving!"

Those who complain about Nazis and other groups they do not like are those who simply do not understand what free speech looks like. Free speech is speech we don't like. Those who intend to use a network that permits them to speak freely must first learn to allow everyone else the same privilege. Those who cannot learn this cannot speak freely, either. They are ultimately destined to return voluntarily to the regular Internet, where they are not subjected to the free speech of others, but where they are also forced to submit to whatever moderation someone else has chosen to impose. Since much of this moderation is ultimately in response to laws passed by governments, or threats of such laws being passed, users of the regular Internet are not free to speak critically of their governments, groups protected by their governments, or topics their governments have decided to censure.

Often, in the whiny voice of an overly-pampered teenage girl, a novice user of an alternative network will complain, "This is too hard to use." To be fair to this category of user, why should someone who simply wants to use a peer-to-peer social network be forced to understand how it functions below the surface? Many developers really need to spend more effort writing software that is usable by this class of user. On the other hand, I have found that while some alternative network software is difficult to use, some is no harder to use than, say, Reddit. The solution here is for new users to recognize that they need to expend the same level of effort learning to use alternative networks as they have learning to use Reddit or Facebook.

Another increasingly common reaction to alternative networks by users of the world wide web is to lump all alternative networks together and see them as being designed solely to make their developers rich through the sale of cryptocurrencies. One commentor on cheapskatesguide.org said in part, "I share your view on NFT, but sadly not your mildly optimistic view on a blockchain-based Web3/Web 3.0. Web3 is thus far funded by VCs (Andreesen Horowitz) and as a commenter noted, seems like 'me-centralization' rather than de-centralization, i.e. recentralize around *me* so I can snipe the money that Meta/Facebook, Alphabet/Google, Amazon are currently making." In accordance with this view, some Internet users see "Web 3.0" as an evil scam. I think part of the reason for this is that they have not asked what "Web 3.0" really means. Perhaps this is because opportunists to have hijacked the term and used it for their own ends, the way IBM did with the term "personal computer" back in the early 1980's. Wikipedia credits Gavin Wood the cofounder of Etherium with the idea behind Web 3.0 back in 2014. However, the term "Web 2.0" was coined by Darcy DiNucci in 1999 and popularized by Tim O'Reilly and Dale Dougherty in 2004. Given the wide-spread usage of the term "Web 2.0" over the decade preceding 2014, I seriously doubt that Wood could claim that he coined the term "Web 3.0" and, therefore, had the right to define it however he chose. Regardless, for whatever reason, many simply condemn all alternative networks under a single label that they do not understand.

Not all alternative networks are supported by the sale of cryptocurrencies. A few use technologies borrowed from cryptocurrencies. For example, ZeroNet relies on the cryptography of bitcoin, but bitcoin is not required to use ZeroNet. ZeroNet is completely free to use, and to my knowledge no one has made a penny off of it, aside from a few voluntary donations to its inventor. I have used ZeroNet daily for three years, and I have never paid a penny, or any fraction of any cryptocurrency, for the privilege of doing so. In addition, I2P, IPFS, Gemini, Gopher, Tor, Secure ScuttleButt, and Aether are not supported by the sale of cryptocurrencies.

Another problem that some potential users have with alternative networks is that, while most do not require the payment of cryptocurrencies for their use, they are not free. An article entitled, Why we need the distributed web describes one positive attribute of the IPFS network. "The distributed web also helps content producers cut down on their distribution costs. The more popular a file is on the distributed web, the less of the bandwidth burden and costs must be covered by the content producer directly. As a content producer, you technically only need to maintain a single IPFS node to ensure long-term availability of your content across the open internet." Reading between the lines reveals the flip-side of the coin. When your computer is a node on the IPFS network, you are paying to host other people's content on your computer. You are paying with your hard drive space and part of your Internet bandwidth. Since IPFS content creators are only paying a small part of the cost of hosting their content, some are likely to be wasteful of the network resources that are being donated by other peers.

Wasting resources that other people are paying for is part of the problem with the current Internet. Once JavaScript allowed website creators to pass much of the computational costs of their websites to users, creators were freed of the burdens of running their bloated inefficient code. This meant they could put much more of it on their websites. Of course, those who understand the history of communism could have predicted the effects of JavaScript on the Internet. Thanks to human beings' unwillingness to work without being compensated in proportion to their labor, communism often ends in near starvation. Those who are capable of leaving, do. Unfortunately, the engineers and computer scientists who write the software that runs the Internet are not known for their prescient predictions of economic outcomes based on their keen understanding of historical principles and human nature. As an engineer who has worked shoulder-to-shoulder with computer scientists and engineers for decades, I am qualified to make that criticism.

Any network--centralized, decentralized, or peer-to-peer--that allows content creators to overly burden users is ultimately doomed to failure. Annalee Newitz, authour of an article entitled, The Internet as we know it is doomed, predicts that the Internet will ultimately follow the example of neolithic cities that disappeared as a result of the wide-spread rise of a wealthy and powerful class roughly 7500 years ago. She surmises that, even back then, many of the wealthy and powerful were acting as societal parasites who ultimately killed their hosts.

On networks like ZeroNet, where each user chooses what he has on his computer and what he is sharing with others, the vast majority of users simply choose not to download sites that contain large quantities of data. This means that when a user goes to one of these sites, most likely, no content exists there because no one is actually hosting (AKA "seeding") it. This is one of the main reasons users leave ZeroNet. Ironically, what causes unaware users to leave, is the very thing that encourages many of us to stay. Having a choice of what ZeroNet content I maintain on my computer is one of my favorite things about ZeroNet, and perhaps one of the top two or three characteristics that has caused me to continue using it. The fact that I can use ZeroNet strictly for accessing efficient personal sites and text-only social media sites means that my computer and Internet connection are not burdened with the serving of thousands of cat videos and inane social media meme pictures. In contrast, I hesitate to use other networks like IPFS, I2P, and Freenet where I have no control over what is occupying my hard drive space and tying up my Internet connection bandwidth.

On networks that developers have set up to essentially tax every user via a cryptocurrency to pay for the costs of running the network, my prediction is that users will simply decide to go elsewhere. They will rationally ask why they should pay to use one alternative network when they can use others for "free". This is one reason I think alternative networks based on cryptocurrencies are ultimately doomed to fail. When it comes to these networks, in my opinion, the "Web 3.0" haters are justified.

A problem closely related to the cost to users of running a peer-to-peer network is the time required to produce the type of quality content that is likely to attract more users. When no corporations use a network, its content is solely up to the users to create. If they are not willing to invest their time in creating high-quality content, the network will not attract more users. This is perhaps the single largest problem with alternative networks. ZeroNet is filled with exhortations by users to their fellows to step up and produce attractive content, but this has largely not happened. As a result, alternative networks are mostly filled with what I call "zombie blogs", where users began blogs, posted two or three articles, and then disappeared. This is a problem that technology cannot solve. Users must respond to this challenge. Unfortunately, I do not see this happening to the extent required to make alternative networks a significant threat to the corporate-dominated Internet, the one possible exception being social media. However, I do not think big social media will feel much of a threat over the next decade or two.

Another problem with even peer-to-peer networks is that he (or they) who controls the code controls the network. This means that for any peer-to-peer network to provide more than false claims of privacy, anonymity, security, and free speech, the code must be under the control of the right person or people. Since money and power corrupt most people, alternative networks will likely rise and fall just as corporations and governments (and apparently neolithic cities) always have. As corporations and governments the world over have repeatedly demonstrated, many leaders are willing to destroy their foundations of power over the long term in order to grab a little more power in the short term. This means that few peer-to-peer networks will operate in "freedom mode" forever. The only check against the de-evolution of a peer-to-peer network into a mini dictatorship or oligarchy is that users have the freedom to leave and join a better network. Therefore, users should plan on using a particular alternative network only as long as it satisfies their needs and then moving on when it ceases to do so.

Now, I would like to mention one more issue before bringing this article to a close.

Cloud Computers will be Inherently Incompatible with the New Internet

I want to explain what I meant when I said the new breed of locked-down, cloud computers will not run peer-to-peer dApps. Over a year ago, when I wrote, Toward a Technological Cage for the Masses and Taking a Stand in the War on General-Purpose Computing, I predicted that personal computers would become more locked down and cloud-based, meaning less like the PC's of the 2010's and more like computer terminals of the 1980's. Since then, Microsoft has created something called the Windows 365 Cloud PC that is a virtual PC that consumers and businesses rent by the month and access over the Internet. It allows a user to run Windows, Microsoft Office 365, Visual Studio, and other Microsoft software in the cloud and access them from his Internet browser. One might speculate that Microsoft offers only a limited number of Microsoft applications on its cloud PC's, and that users cannot upload their own software. However, I could not find a single Cloud PC review that confirmed or denied this. Apparently no reviewer thought this point important enough to clarify. In addition, the Home version of Windows 11 that was released a few months ago cannot even be installed without an Internet connection. Microsoft seems to be slowly conditioning its users to accept cloud computing, and I am continually amazed by the fawning reviewers of Windows products who proclaim, as if with one programmed voice, that this is "the future of computing". Not if I have anything to say about it. I suffered through this in the 1980's, and I did not like it one bit!

When we can no longer run software on computers in our physical possession, we no longer have the option of running the software required to access alternative networks. I cannot help but wonder if the rise of alternative networks and the push for cloud computing are related. Are companies foisting cloud computers on unsuspecting consumers--and make no mistake, they are foisting--because they recognize the threat of alternative networks and are rushing to lock users into the regular Internet that companies control? I don't know. I think more likely, companies are simply salivating at the prospect of locking consumers into paying for additional services, and simultaneously locking them out of alternative networks is just an additional bonus of the cloud computing approach.

Final words

Barring unforeseen developments, the most likely solution to the lack of personal privacy, anonymity, and free speech on today's regular Internet lies with alternative peer-to-peer networks. Although some of the technical problems have not yet been solved, temporary partial solutions exist (including hybrid peer-to-peer networks like ZeroNet, Secure Scuttlebutt, and others) that force governments to engage in costly efforts to spy on and control their citizens' use of the Internet. Since most governments are unwilling or unable to pay for these measures, much of the population of the world currently enjoys access to alternative networks, a least for those who are willing to learn how to use them.

While peer-to-peer networks are the main source of near-term hope for more privacy, anonymity, and free speech on the Internet, they come with some rather severe limitations. They are not suitable for passing around the large quantities of data that most Internet users have grown accustomed to accessing. They are not right for watching hours of Netflix or YouTube videos a day. Nor, are they useful for large social networks with large quantities of social meme videos and pictures. They can be very well suited to text-only communications like world news articles, text-only small and medium-sized social networks, and text-only informational websites. Fortunately, these are exactly the types of digital communication that facilitate free speech and the passing of vital information about the outside world to average citizens living under despotic regimes.

For the reasons mentioned in this article, much of the "new Internet" is likely to look very much like the old Internet for many, perhaps most--only with even less privacy, anonymity, and free speech. But, those who care about such things and are willing to work to have them are likely to be able to find them over the next decade or two on alternative peer-to-peer networks.

If you have found this article worthwhile, please share it on your favorite social media. You will find sharing links at the top of the page.