AI’s false revolution: Study buddy or smoke and mirrors?
Billions spent, minds dulled, and little to show for it – we need to keep control of our thinking in the age of artificial intelligence, writes CQUniversity’s Head of Educational Neuroscience, Professor Ken Purnell.
Billions of dollars are being poured into artificial intelligence systems that promise to change how we learn and work. Yet, according to one major 2025 MIT study, 95 per cent of corporate AI investments have so far generated zero return.
This sobering statistic reveals an uncomfortable truth: the AI revolution isn’t delivering the transformation many expected. Instead, we’re witnessing a widening gap between the dream of a tireless digital study buddy – one that helps us think, write and solve problems – and the reality of fragile, energy-hungry technology that often costs more than it contributes.
Brains on autopilot
In education, AI’s potential is both exciting and perilous. Imagine a world where everyone has a personal study companion – endlessly patient, lightning-fast and always available. Today’s most advanced large language models can already explain complex ideas, edit essays and offer creative inspiration.
But neuroscience tells a different story. When people outsource their thinking to AI, brain connectivity and learning capacity actually weaken. Genuine learning depends on active engagement: questioning, refining and extending what AI suggests.
As I often remind students, you must lead the thinking – AI can only follow.
The real power comes from collaboration, not delegation. Don’t outsource your brain.
Mirage of returns
In the corporate world, the “AI revolution” looks much the same.
Most projects stall before showing measurable results. Only a small fraction – roughly five per cent – achieve lasting impact, and those are the ones that embed AI strategically and focus on targeted, solvable problems.
The rest find themselves caught in what I call the “prestige loop”: investing in AI for status, not strategy. In too many cases, “AI” becomes a convenient scapegoat for layoffs or underperformance. But technology isn’t making these decisions – leaders are. It’s time to question the mantra that “AI did it”.
The hidden environmental bill
Meanwhile, hidden environmental costs are mounting. Every AI prompt draws on vast, energy-hungry data centres that consume staggering amounts of electricity and water. By the late 2020s, AI-specific electricity use could rival a quarter of household power consumption in developed nations, and in some regions, pushing data centres toward 15 per cent of national energy loads. The convenience of instant answers comes with real-world costs and these – costs we no longer can afford to ignore.
A fragile partnership
The most valuable AI isn’t one that runs on autopilot. It’s one used critically as a partner that challenges our thinking, reveals blind spots, and helps us see problems differently. It isn’t about automation – it’s augmentation.
The danger lies in passivity. Overreliance on AI breeds cognitive laziness and dulls expertise. When people accept AI’s output uncritically, they lose the opportunity to think deeply or creatively themselves.
As institutions chase the next big thing, success will belong to those who pair human judgment with technological capability, not those who surrender one to the other. The best outcomes happen when we stay in the driver’s seat – not when we hand over the wheel.
Counting the real costs
Whether in classrooms or boardrooms, the myth of effortless AI progress has collided with hard reality. Real returns are rare, sustainability costs are rising, and unchecked reliance threatens the very skills society needs most: creativity, critical thinking and ethical judgment.
If AI is to be truly revolutionary, it must remain under human command – a critical friend rather than a replacement.
The true measure of progress won’t be how intelligent our machines become, but whether we stay smarter than the tools we create.
So, don’t outsource your brain – lead, question and use AI critically.
