Google is closer to AGI than I am to a billion

Google dropped another banger tool… thats why I think they are close to getting AGI

Google is testing Disco, an experimental Chrome feature built on Gemini 3 that turns open tabs into small interactive web apps called GenTabs, created from plain language prompts and linked sources.

It runs inside Chrome, can reference multiple tabs and past Gemini chats, and for now only works on Mac OS through Google Labs with a limited waiting list.

This feels like Google trying to make browsing less scattered by letting people shape what they already have on screen, instead of asking them to start fresh in a separate AI tool. The lack of required coding matters, because it lowers the bar for turning research or planning into something structured without leaving the browser.

If this works, Chrome becomes more than a place to read pages and starts acting like a workspace where tabs harden into tools.
That could quietly shift how students, planners, and researchers organize their work day, especially for repeat tasks.

The real test is whether people keep these apps open after the screen glow fades.

Explore it here homies Disco by Google

The job market shift people missed

US unemployment rose to 4.6%, up from 4.2% a year ago, but the real break shows up by age. New college graduates are now close to 10% unemployed, while experienced workers remain below the average, according to Fed-linked data Josh Bersin cites.

This looks less like mass layoffs and more like companies freezing the bottom of the funnel while they experiment with AI and control costs.
That choice may be shortsighted, because younger workers tend to adopt new tools faster and question old workflows instead of optimizing them.

At the same time, jobs not requiring a degree now make up about 82% of the workforce, shrinking the traditional college pipeline.
Add in the trust gap, only 27% of US workers say they trust CEOs on AI job claims, and hesitation turns into drag.

If entry-level hiring stays frozen, where does the next wave of builders come from?

AI training work just got pricier

A new HireArt report shows AI model training is shifting away from generalist data labelers toward subject-matter experts with deep domain knowledge and language skills.

As AI models improve, companies want nuanced judgment in fields like law, medicine, and engineering, not low-cost crowd work tagging simple examples.

This is a clear sign that AI development has moved into a more serious phase, where mistakes cost more and shallow inputs break systems in visible ways.
Pay tells the story: generalist trainers earn about $12.50 to $15.50 an hour, while experts can make over $100 an hour, with medical specialists reaching $180.

That gap reshapes who gets pulled into AI work and who gets pushed out, especially as trust and safety depend on decisions made during training.
It also weakens the idea that AI progress runs on endless cheap labor alone.

When the screen glow fades at midnight, whose judgment do you want shaping the model?

Prohuman team

Free Guides

Explore our free guides and products to get into AI and master it.

All of them are free to access and would stay free for you.

Feeling generous?

You know someone who loves breakthroughs as much as you do.

Share The Prohuman it’s how smart people stay one update ahead.