History is littered with confident predictions about technology. Some dismissed innovations that would go on to reshape the world. Others promised revolutions that arrived years later than expected. As AI dominates today's headlines, what can yesterday's mistakes teach us about tomorrow?
In February 1995, Newsweek published an article titled "The Internet? Bah!" In it, astronomer and writer Clifford Stoll argued that online shopping wouldn't replace physical stores, digital newspapers wouldn't overtake print and virtual communities would never become part of everyday life.
Today, those predictions seem almost absurd ~ but that's the point.
Stoll wasn't uninformed. He looked at the technology available at the time and reached what many considered to be a reasonable conclusion. What he couldn't predict was how quickly everything around the internet would evolve. Faster broadband, cheaper computers and later smartphones transformed a slow, niche network into the backbone of the global economy.
History is full of moments like this. Intelligent people, armed with the best information available, making predictions that simply didn't age well.
As artificial intelligence dominates today's conversation, it's worth remembering that technology rarely follows the path we expect.
Looking back, most failed technology predictions fall into one of three categories.
First, we underestimate technologies in their infancy. Early versions are often expensive, unreliable or limited, making it difficult to imagine what they'll become.
Second, we overestimate how quickly breakthroughs will reach the mainstream. Solving a technical challenge is only part of the journey. Regulation, infrastructure, economics and public trust all take time.
Finally, we focus too much on the technology itself and not enough on people. The biggest innovations succeed because people change their behaviour, not simply because better technology exists.
With that in mind, here are five predictions that history got spectacularly wrong.
It's difficult to think of a prediction that has aged more dramatically.
Throughout the early 1990s, many businesses viewed the internet as a novelty. Few expected it to become the foundation for global commerce, entertainment, banking and communication.
The internet itself wasn't the breakthrough. The breakthrough came when millions of people found practical reasons to use it every day. Technology only becomes transformative when it solves real problems.
That remains just as true for AI today.
When Apple launched the first iPhone in 2007, the reaction was mixed.
It lacked features that competitors already offered, while Microsoft's Steve Ballmer openly questioned whether consumers would pay a premium for it. BlackBerry and Nokia appeared untouchable, and physical keyboards were still considered essential for business users.
Less than twenty years later, smartphones have become the primary gateway to the internet for billions of people. Entire industries now exist because mobile computing became ubiquitous.
The lesson wasn't that experts misunderstood smartphones. They underestimated how quickly software, hardware and consumer expectations would evolve together.
Artificial intelligence isn't new. Researchers have spent decades making progress in machine learning, computer vision and natural language processing.
What surprised many wasn't that AI became capable. It was how quickly generative AI moved from research labs into everyday life.
Within months of ChatGPT's launch, businesses were experimenting with AI-assisted coding, customer support, research and content creation. Boardrooms that had barely discussed AI suddenly found it at the centre of their technology strategy.
It's a pattern we recognise from conversations across the market. The discussion has shifted from "Should we use AI?" to "How do we implement it responsibly?"
That's a very different question.
Not every prediction underestimated technology.
Around a decade ago, many believed autonomous vehicles would be commonplace by 2020. While enormous progress has been made, widespread adoption remains further away than many expected.
The engineering challenge proved to be only one part of the equation. Safety, legislation, public confidence and real-world edge cases all slowed adoption.
It's a useful reminder that breakthrough technologies rarely move at the speed of headlines.
When Facebook became Meta in 2021, the metaverse was presented as the next evolution of digital life.
Billions of dollars flowed into virtual reality, digital workspaces and immersive online experiences. Expectations were enormous.
The technology itself hasn't disappeared, but the widespread consumer revolution many predicted hasn't materialised. Instead, virtual and augmented reality have found stronger traction in specialist sectors like engineering, manufacturing, healthcare and defence.
Sometimes technology doesn't fail. It simply finds a different audience.
Every generation believes its technological revolution is different.
In many ways, AI genuinely is.
But history suggests we should be cautious of extreme predictions. Claims that AI will replace every software engineer are no more helpful than claims that it will change nothing. The reality is almost certainly somewhere in between.
What we're already seeing is a shift in the skills organisations value. Technical judgement, systems thinking, security, infrastructure and AI governance are becoming more important, not less. As the technology evolves, so too does the role of the engineers building and managing it.
That's something we've seen reflected in conversations with technology leaders across financial services and the wider engineering market. The question is no longer whether AI will have an impact. It's how organisations adapt without losing the expertise that makes innovation possible.
We have a front-row seat to how these shifts play out in the real world. Every day, we speak with software engineers, infrastructure specialists, cybersecurity experts and technology leaders working at the forefront of financial services and trading. One thing has become increasingly clear: the organisations that are adapting best aren't chasing every headline or betting everything on the latest trend.
They're investing in people with the technical judgement to evaluate new technologies, challenge assumptions and implement them responsibly. History shows that today's boldest predictions won't all come true. The real advantage lies in building teams that can adapt, whatever the future brings.
History doesn't suggest that technology predictions are pointless. It suggests they're usually incomplete.
We consistently underestimate technologies in their infancy and overestimate how quickly they'll reshape the world. The biggest breakthroughs aren't driven by innovation alone. They succeed when technology, infrastructure and human behaviour evolve together.
As AI continues to develop, the same principle applies. Some of today's boldest predictions will prove remarkably accurate. Others will look as outdated as the idea that the internet would never catch on.
The challenge isn't predicting the future with certainty. It's recognising that the organisations best prepared for change are usually the ones that remain curious, adaptable and willing to question even the most confident forecasts.
...