**Web 3.0: Why It Was Never Going to Happen - 7 Dec 2025**

By Kieren Nicolas Lovell

/Introduction/

Over the last few years, just after the Blockchain boom (everything
on a Blockchain) and before "everything needs AI", there was a term
that appeared and attempted to cover all of these emerging
technologies, and that term was "Web 3".

/Web 3: Attack of the Killer Applications/

There was, of course, a reason why this never really took hold. What
we actually got was more of the same, but even more isolated. There
is a reason for this, and in this phlog article I want to discuss
why I think this went so very badly, and why, as a result, we will
end up with an internet that we really will not recognise.

So, let us start by defining what the Web actually is, and where it
came from.

/WWW (What and Why Is the Web)/

Let us be incredibly clear here: the Web is not the internet. The
internet existed far before the Web. The Web is just a single
application that makes use of the internet. However, the premise was
then very, very different.

At the time, in the 1990s, different technologies implemented
different aspects of our internet experience. Email, as we all know,
from a user experience perspective, has remained quite unchanged.
Technically, of course, it has evolved, but for how it is
experienced by most users, it is pretty much the same. Especially
for Outlook users, since they still do not have a real search
function in 2025, just like they did not in 1995.

These implementations fitted certain aspects of personal and
organisational requirements. If you needed real-time chat, then we
had the IRC protocol. It was used for chat around the world to make
new friends, but also used for internal communications within
business.

/Different use cases, same technology/

Now, while Microsoft was playing silly buggers with file types in
the world of MS Office, the world of internet communications was
incredibly streamlined. Email attachments would get through even on
different systems, sometimes through gateway services or an SMTP
relay, but on the whole, it worked.

If you had a super expensive PC and a dual-channel ISDN line, you
could still exchange emails with your best friend in Australia with
an Acorn Archimedes, or with your Canadian buddy with an Amiga 1200.
You could arrange a time to connect together, and sign in to any IRC
server of your choice.

Sharing files was, again, really simple. FTP servers were used for
public software updates, NFS and SMB were used internally, and again
were device agnostic. A Sun Ultra 5 had no problem connecting to an
FTP server in order to exchange files with a Windows user in Boston.

Of course, there were limitations. Most people did not have a T1
connection. ISDN dual-channel was in reach for some, but on the
whole, most people were dial-up. However, system-wise, you could use
whatever you wanted.

However, dial-up did have some advantages...

"Noooooo it did not..."

"Oh, yes it did."

Well, not a technical benefit, but it did have an important impact
on how the technology was used. Since not only download speeds were
limited, there was no assumption that end users were always on.

As a result, developers of services had to take users into
consideration. Websites were incredibly light. Images had to be
optimised for dial-up. They had to be designed to work across all
systems. They had to work in a way that allowed an offline mode, so
you could update all of your correspondence, and then sync back.

This had a really interesting by-product that in today's world we
just do not appreciate as much. It helped us to apply precedence to
our work during work hours, and to focus on the task at hand. We had
alternatives for urgent matters: a phone call, an SMS, a pager, or
even an in-person meeting.

A good way to really see how, at this point, the Web was an enabler
rather than a distraction: just ask anyone who had a Palm or a
Psion organiser. They actually used the device as a primary tool to
organise, not to escape.

Now, I do not want you to think that this phlog is going down the
route of some blogs that say "it was always better in the past".
That is not what I am attempting here. What I am stating is this: a
really low-powered device with just a mail app, an IRC client, and a
separate mobile phone was all you needed to be ready as your mobile
powerhouse.

Because emails were not expected to be answered immediately, work
was easier to classify by importance, and most importantly, the
person with the most expensive equipment did not have the edge.
Unless you were processing some really big datasets, any device that
connected to the internet and had those clients would have been
enough to compete with others, to collaborate with all, and to
create meaningful impact.

Now let us look at where this concept of the Web has deviated from
what was originally envisioned, and where we are now.

(a) Web Is Dead; Long Live Mobile Apps

More and more functionality is now being pushed to mobile
ecosystems, with exclusivity and certain features. If you have ever
met anyone with an iPhone, they will admit, after pressure, that it
is not actually the phone they love... it is more Stockholm
syndrome. They cannot leave the ecosystem.

And do not think Android is any different, especially Samsung users.
The ecosystems are incredibly important, and this goes so against
what we were talking about earlier: system-agnostic protocols, open
and documented.

The result is simple: vendor lock-in. You are not with the vendor
because they are good, you are with the vendor because you simply
cannot leave.

Hotel California was supposed to be a song, not a way of life for
tech companies, but we are very much in this situation.

The same applies to cloud providers. Interoperability between cloud
providers is very surface-level. Again, going on about IRC, but let
us really look. A user could be running Windows with an MSN email
account. A user could have MacOS running CompuServe. An Acorn RiscPC
could connect via Pipex Dial.

They would all use the same protocol on any IRC app of their
choice. No gateways, no middleman, and no account pricing per
instance.

Now, it is this state of vendor lock-in that has created this
strange state of affairs with Web 3.0. People claimed that the
future was decentralisation. In reality, it is a cocaine-fuelled
approach to the complete opposite: centralised providers, with you
having to use multiple services.

In every part of our life, we have seen this. Originally you just
needed Netflix. Now you need Netflix, Disney, Amazon, and Hulu.

You, in your business, use Office, and you implement all of your
controls to lock it down nicely. However, your third party uses
Google Meet. And they lock theirs down nicely. The fix? Guest
accounts in both systems. Another account, another system.

We should be asking, "Wait a god-damn minute..." Why can I not call
a Teams user from a Google account, and bring in an Apple FaceTime
caller?

Were we not supposed to do this with XMPP and Jingle (two
technologies that do text, file transfer, and calling, and are ways
to communicate between different providers)?

Yes. Yes we were. The tech companies did their utter utmost to make
sure this did not work. Because, of course, if it did, a new and
shiny startup could make a product better than theirs.

(b) Where Web 3.0 Walks onto the Stage

And this is exactly the point where Web 3.0 wanders in, wearing a
cheap leather jacket and promising to fix everything. "Do not
worry," it says, "we will decentralise it all. Users will own their
data. Protocols, not platforms. Tokens, not log-ins."

The pitch sounds suspiciously familiar to anyone who remembers dial-
up and IRC. Interoperable, open, nobody owning the whole thing. The
difference is that, back then, nobody had to put the word "decentralised"
in a pitch deck to make it true. It just fell out of how the
technology was built.

Web 3.0 tried to sell decentralisation into a world that had already
decided it liked centralisation very much indeed, thank you. App
stores, walled gardens, "Sign in with $BIGCORP", subscription
bundles, cloud credits, loyalty points. We already lived in Vendor
Lock-In Land, and Web 3.0 turned up with a roadmap that assumed the
opposite.

So the rest of this story is really simple, and really depressing.
We took a stack that used to be open by default, bolted platform
economics and surveillance capitalism on top, and then tried to
glue "decentralisation" back on with NFTs and a whitepaper.

Spoiler alert: it was never going to work. But the reasons why it
was never going to work tell us a lot about where the next version
of the internet is going, and why we might not recognise it when we
get there.