AI will end human work (Opinion)

This one is longer than usual. The kind of post that, when I send it to ChatGPT for review, it says nobody will read. But here it goes.

In recent weeks, interviews with senior tech/AI executives have returned to the same line: “AI will replace jobs” — sometimes with very short timelines and absolute certainty.

I’m not saying the risk to the job market isn’t real. But there’s another side to these interviews:

The message “it will replace people now, cut costs, and boost productivity” is very convenient… because it’s exactly the kind of story investors like.

And the other side (without even covering everything) looks like this:

### 1) OpenAI
There are reported numbers suggesting OpenAI may “burn” around $17B in 2026 (with very high figures also reported for 2025). That’s about $1.4B per month.
At that scale, without revenues keeping up, the hype needs to stay alive.

### 2) Nvidia and Microsoft
Jensen Huang clarified that “invest up to $100B” in OpenAI wasn’t a commitment — it was “up to that amount” and “step by step.”
And there’s reporting that Microsoft pulled back on data center projects in the US and Europe due to short-term oversupply risk.

### 3) Elon Musk and xAI
Tesla: $2B to xAI. xAI: a $20B round. And then the news about acquisition/integration with SpaceX.
My read: it’s also a way to make the project “less of a financial sinkhole” by attaching it to companies that “generate revenue.”

### 4) Apple and Google
They announced a collaboration: Apple’s next generation of foundation models will rely on Gemini + Google’s cloud.
Plain translation: even companies with high margins and strong revenue are pragmatic when “building everything from scratch” is expensive and uncertain.
Meanwhile, Google has other revenue streams and is launching new products to replicate the successful AI startups that already have revenue.

### 5) Data centers and infrastructure
In China there are warnings about underuse / idle capacity.
In the US, beyond adjustments, there’s the physical constraint: power and the grid.

### 6) The productivity paradox
There are signs that people using AI aren’t necessarily working less — often the opposite.
And inside companies you see this: they invest without strategy/processes/training and end up with “nice text” + rework. “Workslop” describes that phenomenon well.

### 7) My opinion (as a user)
AI accelerates, simplifies, and helps — but it still isn’t autonomous in real work: it requires constant validation and corrections. And yes, there’s impact on entry-level jobs/internships in some areas, because many entry tasks are precisely the ones AI already does “well enough.”

In summary:
This is very recent and evolving exponentially, so it’s hard to predict what comes next.
There are real AI risks that tend to get ignored in the rush to be “ahead.”
And AGI (Artificial General Intelligence) — the AI that learns and solves genuinely new problems — still feels like “next year for sure… invest and trust.”