Module 1: What Is AGI?

Let's get something straight first.

The word that matters is "General"

Current AI is narrow. Brilliant, but narrow. AlphaFold predicted the structure of 200 million proteins — a Nobel-worthy achievement — but it can't book you a restaurant. ChatGPT writes poetry and code, but it doesn't understand either. Each AI system is a savant: world-class in its lane, useless outside it.

AGI is the opposite. It's an AI that can do any intellectual task a human can do. Not just pattern-match. Not just autocomplete at superhuman speed. Actually reason. Actually learn something new from a single example and transfer that knowledge to a completely different domain.

Here's the test I use: if you could explain a task to a bright 25-year-old with no specific training and they'd figure it out within a day, AGI should be able to do the same.

Why the distinction matters more than you think

Today, every new AI capability requires a separate development effort. Want an AI that writes code? Train it on code. Want one that reads X-rays? Completely different project, different data, different team, different budget.

AGI collapses all of that. One system. Every intellectual task. That's not an incremental improvement — it's the most significant technology ever created by our species. Full stop.

OpenAI's Sam Altman put their estimated cost of achieving AGI at around $100 billion in compute investment. Anthropic's Dario Amodei called it "the most transformative and potentially dangerous technology in human history." These aren't fringe voices. These are the people building it.

What AGI is NOT

This is where most people go wrong.

AGI ≠ conscious. It might have zero inner experience and still outperform every human at every cognitive task. Consciousness is a philosophical question, not an engineering requirement.

AGI ≠ a robot. It could live entirely as software. No body needed.

AGI ≠ human-like thinking. It might reason in ways we can't even comprehend. AlphaGo's Move 37 — the move that no human would ever play, that stunned a world champion — was a preview of alien intelligence. AGI would be that, at scale, across everything.

AGI ≠ evil. Or good. Intelligence is a capability, not a motivation. What matters is alignment — whether its goals match ours. That's Module 6.

The spectrum you're already on

Don't think of AI as a binary switch. Think of it as a dial:

Narrow AI → Broad AI → AGI → ASI (Superintelligence)

We're somewhere in the Broad AI zone right now. GPT-4 can pass the bar exam, write code, analyse medical images, compose music, and explain quantum physics. That's not narrow. But it still hallucinates, lacks persistent memory, can't reliably plan multi-step tasks, and doesn't truly understand causation.

The gap between where we are and AGI is real — but it's shrinking fast. In 2020, AI couldn't write a coherent paragraph. By 2024, it was passing professional exams. That trajectory should concern you. Not because it's scary, but because most people aren't paying attention.

Why you need to care right now

Even if full AGI is 10 years away (and many serious researchers think it's closer), the approach to AGI disrupts everything. Each step closer changes the economy, changes which jobs exist, changes what skills matter.

You don't need to wait for AGI to arrive to be affected by it. You're being affected right now.

This course gives you the clarity to understand what's coming, the frameworks to think about it, and the practical tools to position yourself. No hype. No doom. Just the sharpest thinking available on the most important technology of your lifetime.

---