People keep asking if AI is a bubble, and if it is, when it will finally burst. That question completely misses the point. AI isn’t a bubble; it’s a technological shift on the scale of the internet or electricity, and you can’t uninvent intelligence.
CEO, GoodData.
Every so-called tech bubble history left a permanent layer of value behind. AI is doing the same, only faster. The organizations (and investors) sitting on the sidelines, waiting for a “pop”, will discover far too late that the world has stopped waiting for them.
The only bubble that exists is expectation
Let’s be direct. There is an AI bubble, but it’s a bubble held up by belief, rather than by any meaningful metric of capability.
Valuations have drifted into the realm of science fiction, assuming breakthroughs in AGI (artificial general intelligence) that simply aren’t here. Or may not even be possible with current models. That’s why hype has become part of the business model.
And so the work that truly matters isn’t happening in grand AGI statements, it’s happening in the plumbing. The analytics, IT infrastructure, data movement, and governance layers that make AI usable in the real world, ensuring that data is trusted and secure, AI outputs are reliable and explainable.
Processes can be audited, all while maintaining clear meaning and context through semantic grounding. That’s precisely why many of the loudest players are still pushing coding assistants and shopping tools: it’s easier to demo than to build a full, enterprise-grade data backbone.
The bubble will deflate, yes. Most startups predicated on pure-play AI won’t survive the next two years. CEOs know this. Investors know this. But the technology will keep pushing forward because it’s already embedded into how we compute, reason, and decide. The genie isn’t going back in the bottle, even if a few valuations and companies go up in smoke.
The US innovates, China imitates, Europe legislates
Equally we can’t just ignore the geopolitical reality. The US is moving at the speed of innovation. China is copying at scale. Europe is regulating. Pretending you can legislate your way into technological relevance is delusional. The AI race is starting to look uncomfortably like the Moon race, you either launch, or you watch.
Regulation matters, but it should protect innovation, not suffocate it. And despite some indications that the EU “regulatory superpower” might be starting to bend towards innovation, it's difficult to say that any one country or bloc has gotten the balance right,
The wrong fear entirely
The broader misconception is that AI tools will replace humans and hollow out industries. This argument resurfaces with every automation wave. AI is changing the value humans bring to the table.
The organizations that thrive won’t be the ones using AI as a cover for headcount cuts. They’ll be the ones using it to make their people more capable, more decisive, and more strategic from top to bottom.
Knowledge work is being rewritten in real time
The speed of change is unprecedented, and adds a new layer of considerations all of its own. For the first time, reasoning, interpretation, and decision support can be generated instantly. We’re automating tasks, yes. But we’re reshaping the very mechanics of knowledge acquisition, comprehension and deployment.
Agentic systems are already analysing, proposing, and taking autonomous actions at an enterprise level. The winners among the employees in those enterprises will be employees who know how to direct the system, question it, and integrate its output into real-world judgement. The winners among the employers are the ones who best harness that drive. Competing with the machine is hopeless. Steering it is priceless.
Fraud detection, supply chain resilience, clinical analysis, the pattern is the same everywhere. AI surfaces the signal; humans apply the meaning. The myth of “AI replacing analysts” evaporates the moment you realize the analyst’s job becomes more important when the noise disappears.
Value, not cost-cutting
If you measure AI purely in minutes saved, you’ve already missed the point. Efficiency is table stakes. The real prizes are new revenue lines, new data products, and new business models that simply couldn’t exist before. At least not at scale, and managed in the same way they can be now. AI isn’t a way to shave down headcount. It’s a way to create things you couldn’t have built at all.
Which is why none of this works if organizations don’t invest in their people. Culture is as important as compute. Teams need the confidence to challenge outputs, experiment, and push the tools beyond their defaults. Companies that build this mindset will be the ones defining the next decade - so start training your people now.
The real danger
The threat isn’t that the AI bubble will burst. It’s that leaders will sit around waiting for it to happen, mistaking caution for strategy while others sprint past them.
AI is now just a baseline. It won’t replace people, but it will utterly redefine what they can achieve. And the organizations that embrace that shift now, with speed, openness, and ambition, will own the future.
You must confirm your public display name before commenting
Please logout and then login again, you will then be prompted to enter your display name.