The IT Industry’s Favourite Hobby: Forgetting 🎈

Every few years the technology industry discovers something “unprecedented”. And every few years we behave as if history started on Tuesday.

We panic.
We inflate.
We proclaim revolutions.
Then we quietly forget.

It is almost charming.

Let us take a short walk down memory lane ... assuming we can remember where it is.

The AI That Already Died (Twice)

Today, AI will change everything.

Again.

Apparently this version is different.

What rarely appears in keynote decks is that we have already had two AI winters.

In the late 1970s and again in the late 1980s and early 1990s, funding collapsed. Expectations outran capability. Grand promises met computational limits. Investors walked away. Researchers quietly rebranded their work in order to survive.

The algorithms did not change much.
The hardware did.
The marketing certainly did.

The field did not die.
The hype did.

And here we are again, announcing a revolution with suspiciously familiar mathematics.

Perhaps this time the scale makes the difference.

But structurally? It feels like the same equations, running on better chips, narrated by louder and better-funded storytellers.

The Internet Was Also “The End of Economics”

In the late 1990s, we were told that profits were obsolete.
Eyeballs mattered.
Traffic mattered.
Burn rate was a feature.

Revenue was optional. Discipline was old-fashioned. Physics was apparently negotiable.

The dot-com bubble inflated on the belief that the old rules no longer applied.

In 2000, it burst.

Companies with no revenue vanished overnight. Pets.com became a case study. Office furniture was auctioned. Domain names returned to the wild.

Infrastructure remained. The internet survived. Reality reasserted itself. And now, years later, we stroll through a Jurassic Park of search engines and portals on Nostalgia Avenue, pretending that none of this ever happened.

The lesson was simple: Technology changes distribution. It does not suspend gravity.

We forgot that too.

2008: IT Collapsed “Just Because”

In 2008, the global financial system cracked.

Banks failed. Credit froze. Entire sectors contracted.

IT, which had confidently declared itself immune thanks to digital superiority, discovered that servers still run inside economies.

Projects were cancelled. Budgets evaporated. Start-ups folded. Engineers discovered that venture capital does not function as an independent energy source.

So much for Lalaland and the world of unicorns.
Turns out Dreamland still runs on cash flow, not fairy dust.
There is, inconveniently, a reality behind the pitch deck.

Technology does not float above macroeconomics.

It participates in it.

We learned it.
Then we quietly returned to believing that this time scale, or cloud, or subscriptions, or AI would make us structurally untouchable.

The Billion-Dollar AI Company That Was Not

More recently, we had a celebrated AI unicorn, Builder.ai (formerly Engineer.ai), promising automated everything.

Impressive demos.
Impressive funding.
Impressive valuation.

Board slides sparkled. Conferences applauded. Analysts nodded gravely.

Except the intelligence turned out to be thousands of human contractors quietly doing the work behind the scenes.

The algorithm scaled.
The humans scaled harder.

When the curtain lifted, the valuation collapsed.

The lesson was not that AI is fake. The lesson was that incentives distort narratives, especially when billions are involved.

We did not retain that one either.

The JavaScript Apocalypse

Let us not forget the “left-pad” incident.

A tiny npm package was removed.
Half the ecosystem stalled.

The world did not end. But we briefly rediscovered fragile dependency chains. For a moment, we almost touched reality. We glimpsed a world where software supply chains matter.
We nearly swallowed the red pill.

Then it turned blue very quickly.

There were calls for maturity. For governance. For resilience.

Then we went back to shipping as if nothing structural had happened.

The Pattern

The pattern rarely changes:

1. Something new appears.
2. We declare a revolution.
3. Capital floods in.
4. Expectations inflate.
5. Panels multiply.
6. Reality intervenes.
7. We quietly rename the same concepts.
8. Repeat.

Microservices were going to eliminate complexity.
They redistributed it and added YAML.

Cloud was going to remove operational thinking.
It centralised it somewhere else and invoiced it monthly.

AI will replace thinking.
It will amplify whatever thinking already exists, including confusion.

Multipliers do not choose direction.
They simply scale whatever is already there.

The Difference With Mature Industries

Railways remember accidents.
Aviation investigates failures for years.
Medicine runs longitudinal studies over decades.

Knowledge accumulates.

In technology, we just pivot vocabulary.

Continuous delivery rediscovered operational discipline.
Platform engineering rediscovered internal tooling and programmatic governance.
AI agents rediscover automation.

Reinvention is not the issue.

Pretending there was nothing before is.

Why This Matters

For Friday Fun, it is amusing. For leadership, it is not.

An industry without memory:

• Overpays for hype.
• Underinvests in foundations.
• Repeats governance mistakes.
• Confuses velocity with leverage.
• Treats every cycle as unprecedented.

In a world obsessed with the next wave, the organisation that remembers the last one has an advantage. Not because it avoids innovation. But because it calibrates expectation.

Institutional memory is a strategic asset.

And here is the uncomfortable truth.

Most engineers are not cynical gamblers.
They carry values.
They prefer stability over spectacle.
They value systems that work over narratives that sparkle.
They want to build things that last.

We rarely even speak about salaries.
Many engineers would already feel satisfied evolving inside a system that does not disguise Ponzi schemes as innovation or recycle hype in ways that resemble money laundering with better branding.

Engineers have a sense of honour.
They do not enjoy selling illusions.
They prefer solving real problems.

If leadership built on those values instead of amplifying volatility, IT could contribute to progress rather than theatre.

Memory is not nostalgia. It is moral infrastructure.

A Cheerful Conclusion

The technology industry does not suffer from disruption. It suffers from amnesia.

The good news? Memory compounds.

If you remember:

• AI winters,
• the dot-com crash,
• 2008,
• dependency fragility,
• valuation theatre,

You do not panic as easily.
You do not inflate as quickly.
You do not collapse as dramatically.

And suddenly, you look strangely calm while everyone else rediscovers gravity and history pretends it started on Tuesday.

Happy Friday.
Do not forget. 😉