AI Boom Meets A Costly Reality
Artificial intelligence is moving forward quickly, but the technology boom is creating some uncomfortable side effects. Zoho founder Sridhar Vembu has been speaking about both sides of this transformation, including the rising cost of computing hardware and the opportunities that AI could still create. Recent reports say memory prices have risen sharply because AI data centres are consuming huge amounts of memory and related infrastructure.
That pressure is slowly reaching ordinary technology buyers as well. Smartphones, laptops, servers, and other electronic devices depend heavily on memory components, while manufacturers are facing higher costs for those parts. Vembu has warned that these changes can make normal business operations more difficult, although he continues to see reasons for optimism around artificial intelligence.
Why Memory Prices Matter Now
Memory might sound like a small component inside a laptop or smartphone, but its importance is much bigger than many consumers realise. Every modern device requires memory to run applications, store temporary information, and handle increasingly complicated software workloads efficiently.
The current AI boom has changed that demand equation considerably. Large AI systems require enormous computing infrastructure, and data centres need substantial quantities of high-performance memory to support those workloads. According to recent reporting, Vembu highlighted memory price increases reaching extremely high levels during the past year.
That creates an unusual situation for the technology industry. Companies building AI infrastructure are competing for some of the same components required by traditional computing businesses. When demand grows faster than supply, manufacturers and customers eventually feel the pressure through higher prices.
AI Can Make Hardware More Expensive
The connection between artificial intelligence and consumer electronics prices is not always obvious. A person purchasing a smartphone is not directly paying for an AI data centre, but the supply chain connects these markets in several different ways.
Memory manufacturers have limited production capacity at any particular moment, and massive demand from data centres can change where those components are allocated. Businesses requiring large quantities may also compete aggressively for available supply, creating additional pressure across the broader technology market.
Vembu recently pointed to this wider effect while discussing Zoho’s own infrastructure costs. He said identical server specifications had become several times more expensive over a relatively short period, with memory prices playing an important role in that increase.
Phones And Laptops Face Pressure
Consumers are likely to notice these developments gradually rather than overnight. A smartphone manufacturer may absorb some additional component costs for a while, but sustained increases eventually become harder to manage without affecting product pricing, specifications, or profit margins.
Laptops face a similar situation because modern machines increasingly include larger memory capacities and faster storage systems. Premium computers can be particularly affected when manufacturers use high-end components that are already under strong demand from enterprise and data-centre customers.
This does not mean every smartphone or laptop will suddenly become unaffordable. Pricing depends on several factors, including manufacturing efficiency, currency movements, competition, component contracts, and regional taxes. Still, higher memory costs create another challenge for manufacturers already dealing with intense price competition.
Vembu Sees Another Side
Despite his concerns about AI’s infrastructure costs, Vembu has not positioned artificial intelligence as something that should simply be rejected. His broader argument has been more focused on efficiency, sensible engineering, and avoiding unnecessary spending.
At an ImagiNxt 2026 discussion, Vembu described today’s AI systems as expensive in terms of money, energy, data-centre resources, and electricity. He also argued that India should look for cheaper and more resource-efficient approaches instead of blindly following the biggest AI infrastructure race.
That distinction matters because the AI debate often becomes too simple. Some people describe AI as the solution to almost every problem, while others treat it as an economic threat. Vembu’s comments suggest that there may be another route where businesses use AI carefully while controlling the resources required to operate it.
Smaller AI Models Could Help
One area receiving increasing attention is the development of smaller and more efficient AI models. Instead of sending every task to the largest available model, companies can use smaller systems for simpler jobs and reserve expensive models for situations where they genuinely provide additional value.
Vembu has previously argued that India should consider smaller AI models and other less energy-intensive approaches instead of trying to compete directly with the largest global language models.
This approach could become more important as AI costs rise. Businesses ultimately need technology that improves productivity without creating a cost structure that becomes impossible to sustain. Smaller models, better software engineering, and efficient AI systems could therefore become increasingly attractive.
The AI Opportunity Remains
The growing cost of hardware does not necessarily cancel out the benefits of artificial intelligence. AI can still help companies automate repetitive work, analyse information faster, improve customer support, and develop new software tools.
The bigger question is whether those benefits are large enough to justify the resources being spent. Vembu has repeatedly questioned the assumption that simply spending more money on AI automatically creates better business results.
He has also suggested that companies should concentrate on practical engineering instead of chasing every expensive AI trend. In his view, creative use of cheaper models can sometimes deliver useful results without requiring enormous computing budgets.
Businesses Need A Different Strategy
For technology companies, the current situation could encourage a shift in priorities. Instead of purchasing the most powerful infrastructure available, businesses may start examining exactly how much computing power each product feature actually needs.
That could influence everything from software architecture to cloud spending. Companies might use lightweight models for routine functions while keeping advanced systems for complex tasks that genuinely need them.
Zoho has itself invested heavily in controlling its infrastructure and data-centre environment. Industry reporting has highlighted the company’s long-running effort to maintain greater control over infrastructure costs and data sovereignty rather than depending entirely on hyperscale cloud providers.
Consumers May Become More Careful
For consumers, higher device prices could eventually change buying habits. People may keep smartphones and laptops for longer periods instead of upgrading whenever a new generation arrives.
Manufacturers could also respond by offering different memory configurations across price categories. Budget models might remain focused on basic specifications, while premium devices could carry higher prices because of stronger component demand.
The situation could also make refurbished devices more attractive. If new hardware becomes noticeably more expensive, buyers may start looking harder at older premium smartphones and laptops that still provide enough performance for everyday requirements.
AI Optimism Needs Some Realism
Being positive about artificial intelligence does not require ignoring its weaknesses. AI consumes enormous computing resources, requires expensive infrastructure, and creates new demands across the semiconductor supply chain.
At the same time, refusing to adopt useful AI technology could leave businesses behind competitors that learn to use it effectively. The more sensible approach may involve understanding where AI creates genuine value and where traditional software can accomplish the same task more cheaply.
Vembu’s comments fit into that middle position. He has questioned the enormous costs surrounding the AI investment race while continuing to explore ways AI can become more efficient and useful. His earlier remarks have also focused on adaptability and practical applications rather than treating AI as an unstoppable replacement for human expertise.
India Has A Unique Opportunity
India could benefit significantly if it develops AI systems that prioritise efficiency rather than simply competing on infrastructure size. The country has a large technology workforce, a growing digital economy, and strong demand for affordable software solutions.
That combination could encourage companies to develop smaller models, specialised AI tools, and software that works effectively with limited computing resources. Such products could eventually have demand beyond India as businesses worldwide search for lower-cost alternatives.
The opportunity becomes even more interesting when AI is combined with India’s existing strengths in software services and engineering. Instead of asking only how large an AI model can become, companies can ask how much useful work can be completed with fewer resources.
Conclusion: AI Progress Should Stay Practical
The rising cost of memory and computing hardware shows that the AI revolution has consequences beyond chatbots and software applications. Phones, laptops, servers, and other devices can all feel pressure when demand for critical components rises sharply. Sridhar Vembu’s position is therefore worth watching because he combines caution about AI’s enormous costs with continued optimism about its long-term potential. The future may not belong only to the biggest AI systems, but also to companies that make artificial intelligence cheaper, smarter, and more efficient. Businesses and consumers should watch these changes carefully while focusing on technology that delivers genuine value.
