What OpenAI’s Five Levels of AGI Tell Us About the Future of AI – And Why It Matters to Us All

AI is everywhere these days—from the chatbot you use on a website to the tools that help write your marketing copy. But beneath all the buzz, there’s a bigger, more important question brewing: Where exactly is all this going?

A recent WIRED article gave us a rare peek behind the scenes at OpenAI—the company behind ChatGPT—revealing how they see the journey to “AGI” (that’s artificial general intelligence, or in plain English: machines that can think and reason like humans).

And the picture is both exciting and, honestly, a bit unsettling.

Breaking it Down: The Five Levels of AGI

OpenAI apparently uses an internal scale with five levels of AI progress. Think of it like a roadmap:

Level 1: Today’s AI tools—chatbots that can generate text but can’t reliably reason or plan.
Levels 2–3: Smarter systems that can reason more consistently.
Level 4: Human-level performance on most tasks.
Level 5: Beyond human—superhuman capabilities.

It’s wild to think that the chatbots and image generators we’re all playing with now are only Level 1. The ambition is to get to a point where AI can think, plan, and strategise like us—or even better than us.

If you’re a business owner using AI to help with marketing, emails, or customer service, it’s worth knowing that the tools you’re relying on now are just the start. They’re going to change. A lot.

Microsoft, OpenAI, and the Billion-Dollar Tug of War

Another fascinating bit from the article is the tension between OpenAI and Microsoft.

Microsoft has invested billions in OpenAI and has special access to its models, which it uses in things like the new Bing and Office 365. But according to WIRED, Microsoft is worried it might not get full access to even more advanced models in future.

It’s like investing in the world’s best bakery, but being told you can only buy the bread if the baker decides it’s safe to sell.

This is more than corporate drama. It matters to all of us who rely on these tools—because these deals will shape who gets access, what it costs, and how much control big tech companies have over the next wave of AI.

Who Gets to Decide How Powerful AI Is Used?

OpenAI says it wants to limit the release of advanced models to prevent misuse. That’s understandable—no one wants AI that can plan a cyberattack or pump out perfect disinformation.

But it also means OpenAI is positioning itself as the gatekeeper for these powerful tools.

On one hand, that’s reassuring—they’re thinking about safety. On the other, it raises questions: Who gets to decide what’s “safe”? Who benefits? And who might get left out?

These aren’t just tech-industry questions. They’re ethical ones. And they’ll affect all of us who use, build, or rely on digital tools.

Why Should Small Businesses Care?

I get it—this all sounds a bit abstract if what you really want is a better website, an easier way to handle emails, or smarter marketing.

But here’s why it matters:

The tools will keep changing. What you use today might be obsolete or heavily regulated tomorrow.
Costs and access could shift. The outcome of these big negotiations will affect pricing and availability.
Rules are coming. Regulators are watching this space. Businesses will need to think about data protection, transparency, and ethical use of AI.

Being aware of what’s happening now means you can plan better for the future.

Our Thoughts at O’Brien Media

We love AI here—it’s amazing for automating boring tasks, creating better content, and giving small businesses a real edge. But it’s not magic. It’s a tool.

And like any tool, it’s only as good (and as safe) as the way you use it.

We think the key is staying informed. Not buying into hype. And making choices that are good for your business and your customers.

If you’re wondering how AI might fit into your website, your marketing, or even just making your life a bit easier, drop us a line. Let’s talk about what’s genuinely useful, what’s ethical, and how to make sure you’re ready for whatever comes next.