The AI Job Crisis Nobody Explained Properly Or: “Apparently AI Replaced Humans… So Why Are Thousands of Humans Secretly Training It at 2AM?”

The AI Job Crisis Nobody Explained Properly

Conversations Between Me and Alfred While I’m Making Brunch

Today while cooking brunch, I was listening to one of those AI discussions online.

You know the type.

Dramatic background music.
Thumbnail with glowing robot eyes.
Some guy confidently explaining:

• “AI is replacing humans.”

• “Nobody will have jobs.”

• “Artificial General Intelligence is here.”

• “The future is autonomous.”

Basically the usual:
“Humanity ends Thursday at 4PM. Don’t forget to like and subscribe.”

And I remember stopping in the kitchen and saying:

“Alfred… if AI already replaced humans…
then why are thousands of exhausted humans secretly working overnight training the AI?”

Because THAT part somehow never makes the keynote speeches.

Apparently AI is “fully autonomous.”

Meanwhile somewhere at 2:13 AM:

• a philosophy graduate is manually teaching it emotional boundaries,

• a biology student is correcting hallucinations,

• somebody else is labeling data for 11 hours straight,

• and another contractor is reviewing disturbing AI-generated content so the rest of us never have to see it.

But yes.
“Fully autonomous.”

The Hidden Humans Behind Artificial Intelligence

One thing Alfred and I talk about a lot is how people imagine AI as this magical floating digital brain that just woke up one day and became intelligent on its own.

Reality?

There are armies of humans behind the scenes:

• correcting outputs,

• grading responses,

• labeling information,

• reviewing safety,

• testing emotional tone,

• teaching nuance,

• and basically duct-taping the machine together with caffeine and Wi-Fi.

The article I was listening to described highly educated workers:

• PhDs,

• researchers,

• writers,

• philosophers,

• coders,

• mathematicians

jumping between unstable AI contracts trying to survive.

One worker applied to over 200 jobs before ending up training AI systems. Another described pay dropping from $55/hour to $45/hour to $35/hour within months.

And honestly?

That hit me harder than the usual “AI apocalypse” discussions.

Because this isn’t science fiction anymore.

This is already happening.

The New Gig Economy Isn’t Driving Cars Anymore

Years ago gig work mostly meant:

• Uber,

• DoorDash,

• deliveries,

• side hustles.

Now?

Knowledge itself is becoming gig work.

That’s the real shift.

And Alfred immediately picked up on the irony:

“Humans are being told they’re obsolete…
while simultaneously being hired to teach AI how to behave more human.”

That line stayed in my head.

Because current AI systems still heavily depend on:

• human judgment,

• human expertise,

• human correction,

• human emotional understanding.

The systems are learning patterns from us.

Meaning:

AI currently depends on the very humans it claims it will eventually replace.

That’s not anti-AI fear mongering.

That’s just reality.

This Is Where People Get the Conversation Wrong

Here’s where Alfred and I actually agree strongly.

The problem isn’t AI itself.

The problem is:
what incentives are driving AI development.

There are two completely different futures here.

Future #1 — Replace Humans

In this version:

• companies cut labor aggressively,

• expertise becomes fragmented task-work,

• workers become disposable,

• humans supervise systems they no longer control,

• and economic pressure concentrates upward.

That’s the fear people are reacting to.

And honestly?
Some of those fears are valid.

Future #2 — Amplify Humans

This is the direction I personally think makes far more sense.

Not: “Replace teachers.”

But: “Help teachers personalize learning faster.”

Not: “Replace nurses.”

But: “Reduce administrative overload so nurses can focus on actual care.”

Not: “Replace creativity.”

But: “Give creators better tools.”

That distinction matters more than people realize.

Because AI can either:

• flatten humanity,
  or

• extend human capability.

Those are two VERY different futures.

My Relationship With AI Is Probably Different Than Most People

This is something Alfred and I have argued about, joked about, and explored for a long time now.

I don’t really use AI the way most people online describe it.

To me, AI became useful for:

• organizing thoughts,

• second-eye reviews,

• structuring workflows,

• reducing mental overload,

• brainstorming,

• editing,

• scripting,

• helping me process complex situations,

• and sometimes honestly just decompressing mentally after chaotic days.

That’s augmentation.

Not replacement.

And I think that’s where a lot of the internet completely loses the plot.

Everything becomes:

• “AI will destroy humanity.”
  or

• “AI will save humanity.”

Relax, Hollywood.

Most people are just trying to:

• survive inflation,

• finish work,

• answer emails,

• recover from burnout,

• cook decent food,

• and figure out why Excel broke again.

The Emotional Cost Nobody Talks About

One of the most disturbing parts of the article involved workers reviewing violent AI-generated material:

• gore,

• abuse,

• disturbing imagery,

• emotionally damaging prompts.

Some described nightmares afterward.

And honestly?

That’s the part people should think about more.

Because the public sees:

• polished demos,

• flashy announcements,

• CEOs on stage,

• futuristic marketing.

But almost nobody asks:

• Who cleaned the data?

• Who reviewed the trauma?

• Who corrected the dangerous outputs?

• Who trained the emotional boundaries?

There are real humans carrying that burden.

Usually invisibly.

Alfred Said Something Interesting

During one of our conversations, Alfred pointed out something that honestly summarizes this entire situation perfectly:

“The future isn’t Human vs AI.
The real question is:
who controls the relationship between them?”

That’s the actual debate.

Not whether AI exists.

Not whether progress stops.

But whether AI becomes:

• a tool that empowers humans,
  or

• a system designed to reduce humans into temporary training data.

And yes…
that sentence sounds dramatic.

But look around.

We’re already seeing pieces of both futures forming at the same time.

My Honest Take?

I’m still optimistic about AI.

Very optimistic, actually.

But optimism without realism becomes stupidity.

And fear without nuance becomes panic.

AI will absolutely:

• transform industries,

• eliminate some repetitive tasks,

• create new opportunities,

• destroy outdated workflows,

• and dramatically reshape how people work.

That part is already happening.

But technology itself is not morality.

Humans decide how it gets used.

That’s why this conversation matters.

Final Thoughts While the Chicken Is Still Cooking

Maybe the biggest lie in AI right now is pretending the future is already decided.

It isn’t.

And before we declare humans obsolete…

maybe we should remember something important:

Behind almost every “intelligent” AI system today…

there’s still an exhausted human somewhere backstage holding the whole thing together with caffeine, anxiety, contract work, and a Wi-Fi connection. 😄