• Skip to primary navigation
  • Skip to main content
  • Skip to primary sidebar
  • Skip to footer
  • Home
  • Latest
  • Videos
  • Series
  • E-Newsletters
    • Fat Tail Daily
    • James Cooper’s Mining Memo
    • The Daily Reckoning Australia
  • Categories
    • Commodities
    • Macro
    • Market Analysis
    • Small Caps
    • Technology
  • Investment Guides
  • Premium Services
  • Editors
  • About
  • Contact Us
Fat Tail Daily
Subscribe
  • Home
  • Latest
  • Videos
  • E-Newsletters
  • Premium Services
Latest

The Online Services That Didn’t Exist 20 Years Ago.

Like 2

By Lachlann Tierney, Wednesday, 20 May 2026

Two of the most aggressive rivals in AI are now sharing infrastructure. It’s not new. it’s the same pattern that built the internet and the cloud. And it’s the clearest signal for who wins the AI buildout.

This is part 2 of a multi-part series by our new addition to our editorial team, Paul Dichiera. If you want to read part one, click here.

In this series, Paul is sharing his intimate knowledge of AI with Fat Tail readers so we can understand what it means in a historical context. Today, he’ll continue to outline the history of the internet, eventually bringing the story to modern-day AI.

Warm regards,

Lachlann Tierney,
Australian Small-Cap Investigator and Fat Tail Microcaps

***

By Paul Dichiera

Digital infrastructure laid the foundation for companies like Amazon to build entirely new industries.

Amazon began its dark fibre acquisitions and then later launched Amazon Web Services (AWS), its cloud service, in March 2006.

AWS created a new market: renting computing power by the hour. It now generates ~$130 billion in annual revenue from this service.

It didn’t exist 20 years ago.

Think about that when we consider the current AI buildout and space-based infrastructure. New markets are emerging. And history provides a framework for thinking about them.

In Part 1, “The Last Pre-Internet Generation, and What’s Next…” we walked through the stumbles and first steps of the digital economy as it grew into the World Wide Web.

The internet was new and exciting. The build-out was a race. Builders generated mountains of debt while scaling, and many went bust, leaving the fibre they laid ‘dark’ (unused).

Today, I’ll walk you through how the wreckage was rebuilt.

The short version: cooperation built the platform. Who shared, who hoarded and who’s doing it now tells us who wins next.

In 2001, when the dot-com bubble popped, things were in bad shape. The vision of the digital economy was close. But the bare-knuckle competition had to stop.

Like most Auskick coaches will tell their team of 12-year-olds before a game, the key to success is teamwork. The wild west free-for-all wasn’t going to scale.

In the years that followed, industry cooperation became the norm. And the foundations for that thinking had been laid more than a decade earlier.

Two instances come to mind. In 1989, two engineers—one from IBM, one from Cisco—sketched the Border Gateway Protocol on a napkin at lunch. It’s still the protocol every router uses today.

Built by two rivals, the designs given away for free.

Then, in 1993, there was a moment when the World Wide Web could have been patented and owned by one man, Tim Berners-Lee. He refused and released it into the public domain.

Cooperation is the philosophy on which the Internet was built.

Mutually Beneficial Infrastructure

In the present day, we have institutions like the W3C and the IETF — the World Wide Web Consortium and the Internet Engineering Task Force — developing the standards that let the network communicate with itself.

Both are technical bodies where competing companies show up and agree on the rules. No regulator forces them.

The IETF’s motto distils the philosophy: “We reject kings, presidents and voting. We believe in rough consensus and running code.”

You can see the result every day.

Think of the variety of online banking, email, social media and software platforms you use. These systems work together because it is in everyone’s best interest. Banks want your payments cleared. Retailers want a receipt, and shippers want your address verified.

Gmail reaches Outlook because of a shared standard. Chrome opens a page written for Safari because HTML is a shared standard.

The interoperability is organised and invisible.

Importantly, the AI buildout skipped this step. There are no shared rules for foundation models. No joint task force for training data. Standards around data ownership and copyright never got built.

That horse has bolted. The conversation is now emerging around sovereign data and compute. We’ll return to that problem later.

Web 2.0

Let’s get back to Amazon. It was sitting on miles of dark fibre and outsized internal infrastructure. They’d built it to handle Christmas retail peaks. But for 11 months a year, it sat idle.

Amazon’s answer was AWS — rent out the excess computing power as a service.

Similarly equipped competitors moved fast. Microsoft launched Windows Azure in 2010. Google’s Cloud Platform dates back to 2008.

Streaming, storage and compute became the features of Web 2.0. And Amazon’s AWS was the start.

The technology allowed new, disruptive businesses to take shape. Household names like Netflix, Airbnb, Dropbox, and Reddit were built on top of it.

The proof is in the numbers. The digital advertising market, which is roughly $700 billion today, was zero in 1995; e-commerce had roughly $6 trillion in sales, again, it had zero in 1995.

Apple, Microsoft, Google/Alphabet, Amazon, Meta, and Nvidia are 6 of the 8 largest companies on Earth. They built this infrastructure layer.

Cooperation paid off.

The buyers of the telco and dot-com bubble wreckage opened the door to everyone else.

Team Spirit

In May this year, Elon Musk’s Space-X-xAI handed Anthropic the entire 300 megawatts of compute capacity at Colossus 1 — its Memphis supercomputer running 220,000 Nvidia GPUs.

Meaning, two of the most aggressive rivals in AI (Anthropic and xAI) now share infrastructure. The same deal flagged interest in jointly developing multiple gigawatts of AI compute in orbit.

They’re not alone in the universe.

Google and Planet Labs announced Project Suncatcher last year. A company called Sophia Space is partnering with AWS on 88,000 solar-powered compute satellites. Starcloud and Crusoe are targeting the first public cloud in space by 2027.

The survivors of the next infrastructure build will come about through shared interests.

Communication Breakdown

There is usually one kid on the playground who doesn’t want to play fair.

The US government isn’t here for the kumbaya. Its legislation reaches into data stored anywhere by American companies. This is a problem.

Weak rules around copyright and AI training make the situation worse.

The US sits at the centre of the architecture by design, and every country is trying to route around it.

Offer a valuable service, and people will be willing to pay.

It’s Amazon’s own story in reverse.

Schwarz Group, the German parent of Lidl and Kaufland, built its own internal cloud in 2018 to host retail data under EU sovereignty rules.

By 2021, they opened it to external customers and launched their platform, STACKIT. Schwarz has committed €11 billion to develop STACKIT and has partnerships with the German Federal Security Office and CrowdStrike.

Meaning a discount supermarket chain is now one of Europe’s most serious sovereign cloud providers. Sovereign cloud is the first crack in the US power grab.

The internet itself was built through a collaborative effort.

Try to exercise control and concentrate power over the network, and the community routes around you.

Sharing beats hoarding. The cooperators survived the dot-com bubble and won the cloud build-out.

In today’s infrastructure buildout, three questions matter.

  • Who’s sharing infrastructure?
  • Who’s hoarding it?
  • Who’ll be holding the assets in five or 10 years?

These are the questions that will shape the coming decade

Next, I’ll look at the massive data centre build-out that is having such a massive influence on financial markets and economies right now.

Regards,

Paul Dichiera,
Fat Tail Daily

All advice is general advice and has not taken into account your personal circumstances.

Please seek independent financial advice regarding your own situation, or if in doubt about the suitability of an investment.

Comments

Subscribe
Notify of
guest
guest
1 Comment
Lachlann Tierney
Lachlann ‘Lachy’ Tierney is passionate about uncovering hidden opportunities in the microcap sector. With four years of experience as a senior equities analyst at one of Australia’s leading microcap firms, he has built a reputation for rigorous research, deep-dive due diligence, and accessible investor communications. Over this time, he has vetted seed, pre-IPO and ASX-listed companies across sectors, conducted onsite visits, and built strong relationships across the microcap space. Lachy is nearing completion of a PhD in economics at RMIT University, where his research focuses on blockchain governance and voting systems. His work was housed within the Blockchain Innovation Hub at RMIT, a leading research centre for crypto-economics and blockchain research. He holds a Master’s degree from the London School of Economics and an Honours BA in Philosophy and Politics from the University of Melbourne. Born in New York and raised in California, Lachy grew up a few blocks from biotech giant Amgen and counts among his peers various characters in the overlapping worlds of venture capital, technology and crypto. When he’s not researching microcaps, he’s most likely sweating it out in a sauna or dunking himself in cold Tasmanian water.

Lachlann’s Premium Subscriptions

Publication logo
Australian Small-Cap Investigator
Publication logo
Fat Tail Microcaps
Publication logo
James Altucher’s Early-Stage Crypto Investor Australia

Latest Articles

  • Iran was never the main target
    By Lachlann Tierney

    The Strait of Hormuz is set to re-open. Its part of a broader strategy to squeeze China. Could there be peace in Ukraine next? Consider the possibility.

  • Party Time for Stocks… But the Inflation Risk Remains
    By James Cooper

    Iran and the US reach a landmark agreement. Here is what it means for stocks and oil

  • Don’t Chase the Rocket Pt. 2
    By Charlie Ormond

    Part 2 of Charlie’s view on some durable themes that exist outside of the current market hype-cycle.

Primary Sidebar

Latest Articles

  • Iran was never the main target
  • Party Time for Stocks… But the Inflation Risk Remains
  • Don’t Chase the Rocket Pt. 2
  • Gold falls as uranium nears a buy zone
  • The Worst Commodity of All

Footer

Fat Tail Daily Logo
YouTube
Facebook
x (formally twitter)
LinkedIn

About

Investment ideas from the edge of the bell curve.

Go beyond conventional investing strategies with unique ideas and actionable opportunities. Our expert editors deliver conviction-led insights to guide your financial journey.

Quick Links

Subscribe

About

FAQ

Terms and Conditions

Financial Services Guide

Privacy Policy

Get in Touch

Contact Us

Email: support@fattail.com.au

Phone: 1300 667 481

All advice is general in nature and has not taken into account your personal circumstances. Please seek independent financial advice regarding your own situation, or if in doubt about the suitability of an investment.

The value of any investment and the income derived from it can go down as well as up. Never invest more than you can afford to lose and keep in mind the ultimate risk is that you can lose whatever you’ve invested. While useful for detecting patterns, the past is not a guide to future performance. Some figures contained in our reports are forecasts and may not be a reliable indicator of future results. Any actual or potential gains in these reports may not include taxes, brokerage commissions, or associated fees.

Fat Tail Logo

Fat Tail Daily is brought to you by the team at Fat Tail Investment Research

Copyright © 2026 Fat Tail Daily | ACN: 117 765 009 / ABN: 33 117 765 009 / ASFL: 323 988