Sometimes We Make Mistakes (Meta’s Cost $80 Billion)
Image: TSD/Chat GPT
How to spend $80 billion learning what developers know (well done Meta).
Finally. Mark Zuckerberg has decided to admit defeat, and that’s a real relief for me as a software developer.
I know it’s unpleasant to see failure (and I see enough of that in my own code reviews), but it’s quite another thing when misfortune happens to Meta.
Throwing money at a bad idea doesn’t make it a good one.
In steps Meta’s The metaverse to rather aptly demonstrate this. As Meta decide to wind down “Horizon Zero Worlds Dawn” this marks the end of a “hope this will work” set of experiments that hit the industry over the last few years.
$80 billion. You know how many USB-C cables that is. At the same time our front end developers are restricted to 24” monitors to save money.
The Real Problem
The metaverse didn’t fail because virtual reality is impossible, it failed because nobody really wanted it yet Meta pressed on in any case. Throw money at an area hoping to win big, while the public are only interested in the likes of Roblox and Fortnite.
It’s turned into classic tech industry behavior.
Build something impressive → assume users will adapt → blame users when they don’t.
I’ve worked on projects where we spent months perfecting an architecture nobody asked for… only for the feature to get quietly removed a quarter later.
Sound familiar? It should. This is a pattern that keeps repeating.
Copying Without Understanding
One of the more brutal criticisms of Meta’s strategy was that they “glommed onto the term ‘metaverse’ without understanding it”.
That hits close to home.
When we need to push a new feature in our app we look at how the competition does it. It’s awful, feels awful and leads to awful results.
We copy success without understanding why it worked.
• Google uses complex interview processes → everyone copies it
• Big Tech builds “metaverse” → everyone pivots
• AI gets hot → suddenly every roadmap says “AI-powered”
This is literally one of the core problems with software development today. Copying the big boys without thinking
This is leading to the classic result. They’re $80 billion in the hole. Write it off, write it down and move on.
You won’t be surprised what they’re moving on to.
Here We Go Again
Now Meta is going all-in on AI.
$115 billion this year alone.
If that number doesn’t make you nervous, it should.
Because I’ve seen this movie before.
Big vision
Huge investment
Vague use cases
“This will change everything”
The difference is… AI actually works.
But here’s the catch.
That doesn’t mean the strategy will.
And AI today feels a lot like the metaverse did in 2021.
Not in capability, AI is clearly more useful, but in how companies are behaving.
Everyone is racing to
• Add AI to everything
• Replace humans where possible
• Build infrastructure at insane scale
And just like the metaverse, there’s a risk we’re solving the wrong problem.
From experience, companies don’t usually ask the real questions. Instead of:
“Should we build this?”
They ask.
“How quickly can we build this?”
I once worked on a project where leadership wanted to delete everything and start again because of something they saw online. No analysis. No context. Just vibes.
That’s not strategy.
That’s panic disguised as innovation.
Jevons Paradox and the AI Illusion
There’s a fascinating idea called Jevons Paradox.
In simple terms:
When something becomes more efficient, we don’t use less of it — we use more.
AI makes developers faster.
So what happens?
We don’t reduce work.
We increase expectations.
More features.
More products.
More systems.
And crucially…
More complexity.
Which means more bugs, more maintenance, and more need for experienced developers to clean up the mess.
Exactly the kind of mess created when you rush into the “next big thing.”
You can read more about this dynamic here
The Developer Perspective
From where I sit, this isn’t surprising at all.
I’ve seen:
Features built and scrapped before release
Entire projects canceled because priorities changed overnight
Teams forced to adopt new tech because “it’s the future”
The metaverse is just that… scaled up.
Massively.
And with better marketing.
I guess they’re going to move to lay off 50% of their staff as well. Perhaps it’s all a joke, a Meta joke.
Conclusion
The metaverse didn’t fail because it was ambitious.
It failed because it wasn’t necessary.
AI won’t fail for the same reason.
But parts of the AI boom will fail, spectacularly, for exactly the same reasons:
Misunderstood user needs
Blind copying
Short-term thinking dressed as vision
And in a few years, we’ll read another headline:
“Sometimes… we get things wrong”
Followed by another pivot.
And another $100 billion or so will be gone.
About The Author
Professional Software Developer “The Secret Developer” can be found on Twitter @TheSDeveloper and regularly publishes articles through Medium.com
The Secret Developer is still waiting for someone to approve their $80 billion budget request for fixing code reviews.