There is a confession to make at the beginning of this story: I have been having a love affair with Jamaica since I was a young boy.
Hellshire Beach was one of my favourite places. There were journeys on the back of my grandfather’s truck, floating in the sea in old car tyres, interfering with chickens in the yard and being chased by goats after irritating them once too often. Lizards were caught and raced. At night, peenie wallies were collected in bottles, their tiny lights creating the sort of magic that no screen could reproduce.
There were hours spent looking through a bedroom window in St Catherine and visits to Guy’s Hill in St Mary. Jamaica seemed enormous then: noisy, colourful, unpredictable and completely alive.
Running alongside that affection was another fascination, one that might initially appear to belong to an entirely different world. It was a fascination with technology.
A Spectrum computer was followed by a Commodore 64 and then a Sega Master System. At about 12 years old, I wrote my first computer game. It was not exactly a technical masterpiece. It involved a stick man and was recorded on an ordinary audio cassette, but it worked. At 15 came access to a shared computer, followed eventually by a career in graphic design and teaching design in my early twenties.
There was also a period working at The Voice newspaper, part of the Gleaner family in Britain. Few experiences matched the excitement of hearing about a story before most other people, laying out the newspaper and then watching the finished pages emerge in print. News was not merely information. It was something physical: composed, arranged, printed, folded and carried into the world.
Life subsequently moved into building surveying, engineering and project management. Technology remained a constant companion. More than 50 websites must have been designed or developed along the way.
Yet, after all those years embracing the next machine, platform or piece of software, Sunday mornings now include a ritual that my younger self would probably have found difficult to understand: going out to buy the newspapers.
Not one paper, usually, but several.
Arriving late can mean finding that they have sold out. Familiar faces appear each Sunday to collect their copies, creating the comforting impression that the old habit remains secure. But another possibility lurks behind that observation. Perhaps shops are simply receiving fewer newspapers. A reduced supply can make a declining market appear remarkably healthy.
That thought captures something about the moment now confronting newspapers, businesses and countries. Familiar institutions remain visible, but the foundations beneath them are moving.
The future usually arrives looking ordinary
Google did not initially feel like a civilisation-changing event.
I first used it shortly after leaving college, around the time the search engine appeared in the late 1990s. It was clean, quick and useful, but it did not arrive with a brass band announcing the reorganisation of human knowledge.
YouTube came later. Then Gmail. Each seemed like another convenient service. Only with hindsight is it possible to see how profoundly these platforms altered communication, advertising, entertainment, education, journalism and memory itself.
There was no agreed day on which the old world ended and the new one began. People simply searched for more things online, sent fewer letters, bought fewer printed maps and gradually stopped remembering telephone numbers.
Artificial intelligence may be entering society in much the same way. The difference is the extraordinary speed of its development and the alarming scale of the claims being made for it.
Some investors describe AI as the beginning of an economic transformation greater than the internet. Critics call it a bubble built upon vast expenditure and uncertain profits. Technology leaders warn that increasingly autonomous systems could replace whole categories of employment. A smaller but influential group believes advanced AI could eventually threaten human survival.
It is difficult to know whether to invest, prepare, celebrate or hide the laptop.
The trillion-dollar question
The argument that AI has become a financial bubble is not without evidence.
The Bank for International Settlements estimates that the five largest technology companies are set to spend more than US$1 trillion on AI-related capital expenditure across 2025 and 2026. The money is pouring into specialist computer chips, enormous data centres, electricity generation and the networks required to connect them.
Some of that expenditure is moving faster than company earnings and available cash flow. Increasing amounts may have to be supported by borrowing.
A separate BIS study concluded that competition between AI companies could generate investment approximately 50 per cent above the socially efficient level under its baseline assumptions. Each company fears being left behind, so everyone builds at once. That is how a technological race can become an investment stampede.
The warning signs are familiar: enormous valuations, extravagant promises and companies attaching “AI” to products that previously managed perfectly well without it. During a gold rush, even an ordinary shovel acquires an ambitious marketing department.
Yet a bubble in AI investment would not prove that artificial intelligence itself is a fraud.
The railway booms of the 19th century produced speculation, bankruptcies and enormous losses. They also left countries with railway networks. The dot-com crash destroyed highly valued companies, but it did not destroy the internet. Instead, the internet continued to spread until it became part of almost every serious business.
AI could follow the same course. Many companies may fail. Investors may lose astonishing sums. Some data centres could become expensive monuments to excessive optimism. The underlying technology could nevertheless remain and permanently alter economic life.
The bubble and the revolution can exist at the same time.
Does the technology really work?
There is now sufficient evidence to conclude that AI is more than an elaborate parlour trick.
Generative AI can draft documents, translate languages, produce computer code, examine records, create images and help customer-service teams manage large volumes of inquiries. Stanford’s AI Index reports that adoption has spread with exceptional speed, even if organisations remain far less prepared to govern the technology than they are to purchase it.
Research organisation METR measures the length of tasks that advanced AI systems can complete independently. Its work indicates that the task-completion “time horizon” of leading systems has been doubling approximately every seven months.
That finding should not be mistaken for proof that a machine can replace an experienced professional. Completing a controlled software task is not the same as managing a complicated building project, advising a vulnerable family or understanding the unspoken expectations of a Jamaican client.
Indeed, METR produced another finding that should puncture some of the excitement. In a study involving experienced open-source software developers, participants using AI initially took approximately 19 per cent longer to complete certain realistic tasks. The developers nevertheless believed that AI had made them faster.
That is an important lesson. AI can improve productivity, but it can also create the sensation of productivity. Producing 20 pages in a few seconds is not especially useful if a human being must spend several hours finding the confident mistakes hidden within them.
AI does not sprinkle intelligence across a badly managed organisation. Sometimes it merely allows confusion to travel faster.
The end of work?
The International Monetary Fund estimates that almost 40 per cent of jobs worldwide are exposed to AI-driven change.
The word “exposed” is important. It does not mean that 40 per cent of workers will lose their jobs. Exposure includes occupations in which AI may automate particular tasks, assist employees or change the skills required.
Routine administrative and entry-level positions appear particularly vulnerable. The same technology that helps an experienced professional complete more work may remove the junior tasks through which a younger employee would once have gained experience.
History offers some reassurance, but not complete comfort. New technology has repeatedly eliminated occupations while creating industries that could not previously have been imagined. The problem is that displaced workers cannot support their families with a historian’s promise that the economy will eventually adjust.
A job may disappear in Spanish Town while the new opportunity appears in California. The statistics might eventually balance; the household budget will not.
The likely outcome is neither the disappearance of all employment nor a painless improvement in productivity. Some jobs will shrink, others will expand, and many will be redesigned. The disruption will be felt most sharply by people who lack the time, money or educational support needed to adapt.
Is humanity really in danger?
There is no proof that existing AI systems are capable of independently ending human civilisation.
Current systems make elementary errors, invent references and misunderstand straightforward instructions. They remain dependent on human-built computers, electricity networks and access permissions.
That does not make the wider risk imaginary.
The 2026 International AI Safety Report brought together more than 100 experts nominated by over 30 countries and international organisations. It found increasing evidence of present-day harms involving fraud, misinformation, cyberattacks, manipulation and unreliable automated decisions. It also concluded that methods for controlling more capable future systems remain incomplete.
The claim that AI will inevitably destroy humanity is not established science. The opposite claim, that it could never present a catastrophic danger, is equally unsupported.
There is an important difference between predicting the end of the world and installing sensible brakes before a machine becomes more powerful.
Jamaica cannot remain merely a customer
Jamaica is unlikely to build the world’s largest general-purpose AI model, nor does it need to. The greater opportunity lies in applying the technology to Jamaican problems.
AI could assist teachers, improve access to public information, support medical screening, identify construction defects, strengthen disaster planning and help farmers analyse weather, soil and crop conditions. Small companies could gain capabilities that were once available only to large corporations.
Property and construction provide obvious examples. AI can already help organise listing information, analyse market patterns and produce preliminary designs. But it can also generate fraudulent advertisements, fabricate property images and present dangerously inaccurate advice with convincing authority.
The central question is ownership. If Jamaica becomes only a subscriber to foreign AI systems, the country may import the disruption while exporting its data, expenditure and intellectual value.
Local capability matters. Jamaican universities, businesses and public institutions need people who can build, test, question and govern these systems. Digital literacy can no longer mean knowing how to open an application. It must include knowing when the application is wrong.
What remains after the machine changes?
For several years, a market stall selling CDs helped finance my education. Those little silver discs contributed towards more than one degree.
Do I miss CDs? Not particularly. Some remain in the house, but buying another is unlikely.
Vinyl is different. There are records connected to memories of my grandmother, and music occupies a special place in a family that includes Gregory Isaacs as a second cousin. The slight crackle before a record begins carries a warmth that perfect digital reproduction somehow struggles to capture.
Technology rarely removes everything it replaces. Sometimes the old object survives because it offers an experience the new one cannot.
Newspapers may endure for similar reasons. They provide limits in a world of endless scrolling, a physical record of what mattered on a particular day and the quiet pleasure of turning a page without being watched by an algorithm.
Artificial intelligence may also surprise both its evangelists and its critics. Governments could slow its development. Technical limitations could prove harder to overcome than anticipated. The investment bubble could burst. Or AI could continue advancing until it becomes as ordinary and indispensable as Google, quietly woven into almost every part of life.
The Future Never Announces Itself
The evidence points towards a verdict that lacks the simplicity demanded by both sides of the argument.
AI is probably a bubble. It is also probably a revolution.
The inflated promises will not all survive. Many businesses will fail and considerable money will be lost. But the underlying technology is unlikely to disappear. It will change professions, institutions and personal habits, often before society fully understands what has happened.
The future has done this before. It does not knock on the door, introduce itself and wait politely to be invited inside.
It arrives quietly, looking like a search box, a cassette containing a stick-man game, or a strange new tool that does not yet seem capable of changing very much.
Follow Jamaica Homes on Youtube @jamaicahomes and Instagram @jamaica_homes and on Facebook @jamaicahomesnews Send us a message or email us at onlinefeedback@jamaica-homes.com or editor@jamaica-homes.com
Visit our YouTube Community ↗