Why the gap between what AI can do and what most people believe is now dangerously large — and what you can do about it.

The organisations that invest in AI literacy and development now, while competitors are still in the "this seems overblown" phase, will define the next era. The ones that wait will spend twice as long closing a gap that keeps widening.

What sparked this

Matt Schumer's viral X post "Something Big is Happening" (84M views) stuck a nerve because it articulated something many people working closely with AI had already been feeling: the world hasn't caught up to where this technology already is.

The full article is 5,000 words; I pulled out the key points and added what I think is missing.

We will cover:

Something Big is Happening.

The post draws a striking parallel to February 2020, weeks before COVID reshaped everything. Most people dismissed the early warnings then, too.

The pattern is repeating. But there's one critical difference this time: preparation actually works. You don't need to wait for a vaccine. The tools are already here, available today, ready to be used.

These last few weeks were intense.

New model releases. Claude Cowork, OpenClaw levelled up the AI game. I went into builder mode, and set up OpenClaw. Working with these tools wasn't plug-and-play. They required setup: capturing workflows, giving AI context, granting access to files and business insights. But once it clicked, once the AI actually understood how the work flows and could run repeatable processes, the entire day-to-day changed.

I had a glimpse of the future, and it was … wow… and a bit scary.

The feeling wasn't fear that AI is "coming for us." It was something more nuanced: realising that the distance between people who are building with these tools and people who haven't started yet is growing faster than anyone expected.

Every week the gap widens. Every month the tools get more capable. And most people are still standing on the sideline, judging AI by experiences that are already outdated.

The models are further along than you think

Four years ago, AI struggled with basic maths. Today, leading models independently build, test, and ship complete software applications. They reason through multi-step problems, demonstrate judgement about quality, and operate autonomously through entire workflows. This is not gradual improvement. It's an exponential leap.

And most people are judging AI on outdated experience. If you tried a free chatbot in 2023 and concluded it wasn't impressive, that's like using a brick-sized phone from 1998 and deciding phones would never be useful for browsing the internet. The gap between a free 2023 model and a premium 2026 model is not incremental — it's transformational.

I see this in boardrooms constantly. Leaders making strategic decisions based on a version of AI that should no longer exists.

They're planning for a world that has already changed underneath them, using reference points that expired months ago.

Why AI labs focused on coding first

There's a reason the labs prioritised making AI excellent at software development before anything else: so AI could help build the next version of AI.

This isn't theoretical. OpenAI has confirmed that GPT-5.3 Codex helped debug and manage the code used in its own creation. AI is now building AI. Each generation arrives faster and more capable than the last, because the tools making it are getting smarter too.

This creates a compounding loop that explains why the pace isn't slowing down. It's accelerating. Each new model makes the development of its successor faster and more efficient. The gap between releases is shrinking while the capability jumps are growing. Understanding this dynamic is crucial for grasping why "waiting to see how this plays out" is a more costly strategy than it appears.

Why you can't just outrun this one

Previous waves of automation followed a predictable pattern: machines replaced specific physical skills, and workers could retrain into cognitive work that machines couldn't reach. That playbook worked for generations. Factory workers became office workers. Manual labour shifted to knowledge work. There was always a safe zone to move into.

AI breaks that playbook entirely. It's a general substitute for cognitive work, and it gets better at whatever new field you try to retrain into. The escape route that worked in every previous technological shift simply doesn't exist in the same way this time.

This doesn't mean careers are pointless. It means the strategy changes fundamentally. Instead of running away from AI into new fields, the winning move is to run towards it. The value shifts from "what you know" to how effectively you can direct and leverage AI as a collaborator. The people who thrive will be the ones who learn to work with AI, not in spite of it.

What the hype misses

Some of Schumer's piece is one-sided, and I agree with it directionally.

It's important to hold two truths at once. AI capability is advancing at an extraordinary pace, and that advance is real and significant. But capability is only one part of the equation. The real bottlenecks are moving at a very different speed.

Implementation friction, data quality, infrastructure gaps, human preferences, and security concerns all create real drag on adoption. An AI model might be capable of transforming a workflow in theory, but the organisation might lack the clean data, the security posture, or the cultural readiness to actually deploy it. Progress isn't just about what AI can do. It's about what organisations can actually absorb.

This nuance matters because it shapes how you should think about preparation. You don't need to panic. But you also can't afford to dismiss the trajectory. The right response is informed urgency: understanding that the technology is moving fast while recognising that thoughtful, strategic adoption will outperform both blind hype and wilful ignorance.

So what do you actually do?

If you can use paid tools, choose one and commit. Right now I'm loving Claude most — it lets you capture repeatable processes as Skills, which means your workflow compounds over time. Gemini is also excellent. Pick one.

Then use Thundamental’s Crawl, Walk, Run, Fly framework:

Start with Crawl means just start. Stop waiting for the perfect moment, the perfect prompt, the perfect use case. Just get familiar. It doesn't matter how well you use it yet. Open it, use it for real work, get comfortable, explore.

It means you push yourself to understand what every button in your chosen tool actually does. Explore the features. Most people use 10% of what's available.

Keep ownership of your work

Here's the line that separates replaceable from indispensable: if your work gets done by pressing one button and you put nothing of yourself into it, you are very replaceable. That's being a Passenger. AI is the Pilot and you go wherever AI takes you. No thought applied. No judgement. No craft.

But if you bring thought, if you elevate what AI produces, if you direct, refine, and apply your own judgement, you bring enormous value. You're the Pilot. AI is your Co-Pilot, but you stay in control and responsible for the outcome.

The bottleneck is no longer technical skill. It's the person with the idea, the taste, the vision, and the drive to put these tools to work. That's where the opportunity sits right now. And it's wide open for anyone willing to step into it.

The Opportunity Is Wide Open

The organisations that invest in AI literacy and development now, while competitors are still in the "this seems overblown" phase, will define the next era. The ones that wait will spend twice as long closing a gap that keeps widening. This isn't speculation. It's the pattern that plays out every time a transformative technology reaches this inflection point.

For individuals, the same logic applies. AI doesn't replace ambition, creativity, or judgement. It amplifies them. The people who learn to work with AI as a collaborator will find themselves with capabilities that would have required entire teams just two years ago.

The window won't stay open forever. But right now, in this moment, the tools are accessible, the learning curve is manageable, and the competitive advantage is real. The question isn't whether AI will reshape your work. It's whether you'll be ready when it does.

Want to explore how this could work in your organisation?

The question is: are you brave enough to try an approach that looks slower on paper but delivers exponentially better results in practice?

👉 Schedule a free consultation and let’s get started.

If you forget everything else, remember this…