Apple and Meta are exploring a future where powerful artificial intelligence runs on personal computers, raising new questions about the billions being invested in data centres and what the changing technology could mean for Jamaica.
KINGSTON, Jamaica, 24 October 2026: A significant shift in artificial intelligence is gathering pace as technology giants Apple and Meta advance systems capable of running sophisticated AI directly on personal computers, potentially changing the future demand for enormous data centres and the land, electricity and infrastructure required to support them.
The development raises questions about the long-term direction of one of the world’s fastest-growing industries. With technology companies spending hundreds of billions of dollars expanding their computing infrastructure, advances in smaller, more efficient AI models could eventually reduce the need for some services to depend entirely on remote servers.
The emerging trend was examined in the BBC’s AI Decoded programme on October 24, which explored how Apple’s hardware and Meta’s approach to AI development could make powerful computing more accessible outside traditional cloud environments.
For Jamaica, the implications extend beyond technology. They touch on electricity costs, commercial development, digital infrastructure and the ability of smaller businesses to compete in an increasingly automated economy.
The Computer Is Becoming the Data Centre
For years, artificial intelligence has largely depended on powerful computers housed in specialised facilities. When a person asks an AI assistant a question, information is typically transmitted to a remote server, processed and returned through an internet connection.
These facilities require substantial physical infrastructure, including land, electricity, cooling systems, telecommunications networks and backup power.
However, advances in computer processors and AI software are allowing increasingly capable models to operate locally.
Instead of sending every request to a distant data centre, businesses could eventually perform more complex tasks using computers inside their offices. Documents could be examined, customer enquiries processed and business information analysed without every operation requiring a cloud connection.
The technology also offers potential advantages for privacy and operational independence, particularly where internet connectivity is unreliable or organisations wish to retain greater control over sensitive information.
That does not mean businesses would no longer need the internet or remote computing services. Rather, the balance between local processing and centralised infrastructure could change.
Apple and Meta Take Different Approaches
Apple’s June 2026 announcement of its third-generation foundation models demonstrated how far local artificial intelligence has progressed.
The company’s new range includes a model containing 20 billion parameters, designed to use advanced computing techniques that activate only a portion of its capabilities during individual tasks. The approach makes it possible to run increasingly sophisticated AI on suitably powerful personal hardware.
Yet Apple’s strategy does not abandon the cloud. The company has also expanded its server-based AI capabilities, including arrangements involving Google Cloud and NVIDIA technology for more demanding applications.
Meta has pursued another route by supporting AI models that developers and businesses can deploy on their own computing infrastructure.
Such developments could allow organisations to exercise greater control over their AI systems rather than depending exclusively on services operated by major technology providers.
There is, however, a striking contradiction.
In July 2026, Meta announced construction of a major AI data centre in Alberta, Canada, representing an investment of more than CAD $13 billion. The facility is designed to have a power capacity of one gigawatt.
The company is therefore helping expand access to independently operated AI while continuing to invest in enormous centralised facilities.
Both developments can exist together. Smaller systems may become capable of handling everyday work, while larger facilities remain necessary for training advanced models and supporting computing-intensive services.
The Growing Pressure on Land and Electricity
Despite improvements in AI efficiency, international forecasts suggest that global data centre demand will continue rising.
According to the International Energy Agency, data centres consumed approximately 485 terawatt-hours of electricity worldwide in 2025.
That figure is projected to reach around 950 terawatt-hours by 2030, nearly doubling within five years.
The agency also found that electricity consumption from AI-focused data centres increased by approximately 50 per cent during 2025.
These figures illustrate the scale of the challenge facing governments, energy providers and property developers.
Large computing facilities are not simply technology projects. They are major physical developments requiring suitable locations, reliable electricity, substantial investment and supporting infrastructure.
In some countries, the expansion of data centres has prompted debate about pressure on electricity networks, water resources and competing land uses.
Improvements in local computing could eventually change the type and scale of facilities required. However, current evidence does not indicate that demand for large data centres is approaching an end.
What Could This Mean for Jamaica?
Jamaica’s position is different from countries hosting enormous concentrations of global computing infrastructure.
The more immediate significance lies in how businesses, institutions and households could use AI without depending so heavily on expensive remote computing services.
For property professionals, construction firms and small businesses, locally operated AI could support administrative work, document processing, customer communication and analysis of business records.
A real estate company, for example, might eventually use an office-based AI system to examine property information, organise listings or identify inconsistencies in records while keeping more of its internal data under direct control.
For developers and investors, the wider shift also raises questions about future commercial property requirements.
If more computing takes place within ordinary offices, the demand for some specialised computing services could change. At the same time, growing digital activity may create opportunities for telecommunications facilities, smaller regional data centres and buildings designed to accommodate more advanced technology.
There are limits to these possibilities. Powerful computing equipment still requires capital investment, electricity, maintenance and technical expertise.
In Jamaica, where energy costs and infrastructure reliability are important considerations for commercial operations, those practical requirements would remain central to any decision to adopt local AI.
There is currently insufficient evidence to suggest that decentralised artificial intelligence will substantially reduce Jamaica’s future electricity demand or alter its property market in the near term.
A Different Future for Digital Infrastructure
The emerging development is less about the disappearance of data centres and more about where computing power will be located.
The industry appears to be moving towards a combination of powerful personal devices, privately operated business systems and enormous centralised facilities.
Each serves a different purpose, and the relative importance of those systems may change as hardware becomes more capable and AI software becomes more efficient.
For governments and investors, the challenge will be understanding how quickly that balance develops before committing substantial resources to infrastructure designed to last for decades.
For Jamaica, the opportunity may lie in making advanced computing more accessible to businesses while ensuring that electricity, telecommunications and commercial infrastructure can support changing technological demands.
The buildings that power artificial intelligence are unlikely to disappear. But the next generation of AI may increasingly operate inside the buildings where people already live and work.
And that could prove just as significant as the construction of another giant data centre.
Visit our YouTube Community ↗