The global shortage of computer memory has become one of the biggest technology discussions of 2026. RAM prices are rising, PC builders are finding it more expensive to upgrade their systems, and manufacturers are facing increasing pressure to secure enough memory chips. At the centre of this debate is the rapid expansion of artificial intelligence and the growing demand for powerful data centres.
Now, OpenAI CEO Sam Altman has added another dimension to the discussion with his controversial comments about how the world’s limited RAM supply should be used. According to Altman, keeping large amounts of memory inside personal computers is inefficient when that memory remains unused for much of the day. Instead, he believes that cloud data centres can make better use of the same resources by sharing computing capacity among multiple users.
His comments have attracted attention from PC enthusiasts, gamers, developers and technology professionals. Many people believe that users should have the freedom to own powerful computers and run AI models locally. Others see his argument as a practical approach to managing limited global resources.
But what exactly did Sam Altman say about RAM? Why are memory prices increasing? And could the growing demand for AI data centres change the future of personal computing?
What Did Sam Altman Say About RAM?
Sam Altman discussed the global memory shortage during an appearance on the Cheeky Pint podcast, hosted by Stripe co-founder John Collison. The episode was published on October 29, 2026, and covered several topics related to artificial intelligence, computing infrastructure and the future of technology.
During the conversation, Altman questioned the idea that more people running AI models on their personal computers would solve the growing demand for data-centre infrastructure.
He argued that many computers contain large amounts of DRAM that are not being fully utilised throughout the day. From a global resource-allocation perspective, he described this as a poor use of the world’s limited memory supply.
Altman suggested that memory could be used more efficiently in cloud data centres, where computing resources are shared across multiple users and applications. He estimated that personal computers use their available DRAM only around five per cent of the time.
This figure was his conversational estimate, not a universal measurement of how every computer uses memory.
His central argument was simple: if the world is facing a shortage of memory chips, why should large amounts of RAM remain installed in individual computers when cloud infrastructure could potentially use that capacity more intensively?
However, the statement quickly became controversial because it came from the CEO of a company that is investing heavily in AI infrastructure.
Why Are Sam Altman’s RAM Comments Controversial?
The biggest reason behind the criticism is the growing difference between what AI companies need and what ordinary computer users want.
For years, computer manufacturers encouraged consumers to purchase laptops and desktops with more RAM. Higher memory capacity became an important selling point for gaming, video editing, software development, multitasking and professional applications.
Today, many users also want powerful computers capable of running artificial intelligence models locally. Running AI software directly on a personal computer can offer greater control over data, reduce dependence on online services and provide access to certain tools without paying monthly cloud subscriptions.
Altman’s comments appear to challenge this direction. Instead of encouraging individuals to own more computing resources, he believes that shared cloud infrastructure can make better use of limited memory supplies.
This has raised questions about whether the future of computing will increasingly depend on large technology companies rather than individual hardware ownership.
There is also an obvious conflict of interest in the discussion. OpenAI operates cloud-based AI services that depend on enormous amounts of computing power. A world that relies more heavily on cloud AI could benefit companies that own and operate large data centres.
That does not automatically make Altman’s argument incorrect. However, it explains why many consumers are questioning the suggestion that their personal computers are an inefficient use of memory.
What Is RAM and Why Is It Important?
RAM stands for Random Access Memory. It is one of the most important components in computers, smartphones, tablets and many other electronic devices.
Unlike long-term storage, such as an SSD or hard drive, RAM temporarily holds information that a device needs while running applications. It allows the processor to access frequently used data quickly, helping software operate smoothly.
For example, when someone opens a web browser, edits a video, plays a game or runs multiple applications simultaneously, the computer uses RAM to manage the information required by those tasks.
The amount of RAM available can affect how many applications a computer can handle at the same time. If a system does not have enough memory, it may become slower, particularly when running demanding software.
RAM is also important for artificial intelligence. Large AI models require substantial memory to process information and generate responses. Depending on the model and hardware, running AI locally can require significant amounts of system memory or specialised high-bandwidth memory.
This is one reason why memory has become such an important part of the global technology supply chain.
How AI Data Centres Are Increasing RAM Demand
Artificial intelligence has changed the way technology companies use computing infrastructure. Services such as AI chatbots, image generators, coding assistants and automated agents require powerful processors and large amounts of memory.
Every time a user sends a request to a cloud-based AI service, data centres handle the computing work required to generate a response. When millions of users access these services, the infrastructure needs become enormous.
Modern AI data centres use specialised processors, including GPUs and AI accelerators, along with high-performance memory systems. High-Bandwidth Memory (HBM), in particular, plays an important role in supporting demanding AI workloads.
The growing demand for these components has placed additional pressure on the semiconductor industry.
Memory manufacturers must decide how to allocate production capacity across different products and customers. AI infrastructure has become an increasingly important market, encouraging suppliers to focus on high-value memory products used in servers and AI systems.
This can affect the availability of memory products used in conventional computers and consumer electronics.
OpenAI is also investing heavily in computing infrastructure through projects such as Stargate. Large-scale infrastructure plans from companies including Microsoft, Google, Amazon and Meta are adding to the demand for advanced chips, memory and data-centre capacity.
Consequently, the AI boom is not just changing software development. It is also influencing the manufacturing and pricing of the physical components used in everyday devices.
Why Are RAM Prices Increasing in 2026?
The increasing demand for AI infrastructure is one of the major factors behind the memory market’s current pressure. However, the shortage is not caused by a single company or one technology project.
Several factors are contributing to the situation.
1. Growing demand from AI companies
AI data centres require large quantities of advanced memory. As technology companies expand their AI services, they compete for limited semiconductor manufacturing capacity.
2. Increased demand for high-bandwidth memory
AI accelerators depend on specialised memory products that can transfer large amounts of data quickly. Manufacturers are allocating resources to these products because of their importance in AI infrastructure.
3. Limited manufacturing capacity
Building new semiconductor manufacturing facilities requires significant investment, advanced equipment, technical expertise and time. Memory production cannot increase immediately whenever demand rises.
4. Cautious investment by manufacturers
The memory industry has experienced several periods of oversupply and falling prices. Manufacturers therefore have to balance current demand against the possibility that market conditions could change in the future.
5. Competition between technology companies
Major cloud providers and AI companies are competing to secure memory, processors and other infrastructure components. Long-term supply arrangements can make it more difficult for smaller buyers to access the same resources.
These factors are contributing to higher memory costs, which can eventually affect the prices of laptops, desktops, smartphones and other electronic products.
Cloud Computing vs Local Computing: What Is the Difference?
The disagreement surrounding Altman’s comments becomes easier to understand when we compare cloud computing with local computing.
In local computing, applications run directly on a user’s device. The computer uses its own processor, RAM and storage to complete tasks.
For example, someone editing a video on a desktop computer uses the machine’s internal hardware to process the footage. Similarly, a developer running a local AI model uses the computer’s available resources to generate responses.
Cloud computing works differently. The processing takes place on remote servers operated by a technology company or cloud provider. Users access those resources through an internet connection.
| Feature | Local Computing | Cloud Computing |
|---|---|---|
| Processing | On the personal device | On remote servers |
| Internet requirement | Not always required | Usually required |
| Hardware ownership | User owns the computing hardware | Provider owns the infrastructure |
| Privacy control | More direct control over local data | Depends on provider policies and settings |
| Performance | Depends on personal hardware | Depends on server capacity and connection |
| Cost | Hardware purchase and maintenance | Subscription or usage-based charges |
| Availability | Can work offline | Internet and service availability matter |
Altman’s argument focuses on resource utilisation. A powerful data centre can serve many customers instead of keeping computing resources dedicated to one individual.
However, cloud computing also creates dependence on internet connectivity, service providers and subscription costs. Local computing offers independence and can be particularly useful when users need offline access, privacy or specialised hardware performance.
Why PC Gamers and Developers Disagree With Sam Altman
For PC enthusiasts, RAM is not simply a resource that sits unused. It is an essential part of a computer that allows users to perform demanding tasks whenever they need to.
Gamers, for example, often purchase systems with 32GB or 64GB of RAM to support modern games, streaming, background applications and other demanding workloads.
Developers may require additional memory to run virtual machines, development environments, databases and local AI models. Video editors and designers also depend on sufficient RAM to handle large projects.
These users do not necessarily need their computers to operate at maximum capacity every minute. They need the hardware to be available whenever a demanding task begins.
There are also privacy concerns. Running AI models locally can allow users to process certain information without sending every prompt or file to an external cloud service.
For people working with sensitive documents, private research or confidential business information, local processing can be an important advantage.
Another concern is the possibility of becoming dependent on a handful of large technology companies. If AI computing moves primarily to cloud platforms, users may have less control over pricing, access, service availability and the models they can use.
Therefore, critics argue that the solution to the RAM shortage should be increasing manufacturing capacity rather than discouraging people from owning memory in their personal computers.
Does Sam Altman Want Everyone to Stop Using RAM?
No. His comments were focused on the efficiency of allocating limited global memory resources, particularly when comparing personal computers with cloud data centres.
He did not announce a policy requiring people to remove RAM from their computers or suggest that personal computers should no longer have memory.
His argument was that memory used only occasionally in individual machines could potentially be utilised more intensively through shared cloud infrastructure.
It is also worth considering that Altman discussed the broader future of local AI and cloud computing, rather than making a direct recommendation about the specifications of every consumer computer.
Personal computers will continue to need RAM for everyday tasks, professional applications, gaming and local AI processing. The debate is about how the world’s limited semiconductor resources should be allocated as demand continues to grow.
Will AI Make Personal Computers Less Important?
Artificial intelligence may change the role of personal computers, but it is unlikely to make them irrelevant.
Cloud-based AI services are becoming increasingly capable. They allow users to access powerful models without purchasing expensive hardware. This makes advanced AI tools available to people who do not own high-performance computers.
However, local computing remains important for several reasons.
First, users need computers to interact with software, manage files, create content and complete everyday tasks. Even when AI processing takes place in the cloud, a personal device remains the primary interface.
Second, local AI can offer greater control over personal information. Users can run compatible models directly on their machines without depending entirely on external servers.
Third, internet connectivity is not always reliable. Cloud-based services may become unavailable during network disruptions, while locally installed software can continue working.
Finally, some professional workloads require local hardware for performance, compatibility and operational reasons.
The future is therefore more likely to involve a combination of local and cloud computing rather than the complete replacement of personal computers by data centres.
What Should Consumers Do During the RAM Shortage?
With memory prices under pressure, consumers should think carefully before purchasing a new computer or upgrading an existing one.
If someone already owns a computer that handles everyday tasks comfortably, there may be no immediate reason to upgrade its RAM.
Before purchasing a new laptop or desktop, buyers should consider the type of work they perform. Basic office work, browsing and media consumption generally require less memory than video editing, gaming, software development and local AI workloads.
It is also important to check whether a laptop supports future memory upgrades. Some modern devices use soldered memory, which cannot be replaced or expanded easily.
Consumers planning to run AI models locally should research the memory requirements of their preferred software before purchasing hardware.
Most importantly, buyers should avoid paying for excessive specifications they will never use. At the same time, purchasing a computer with insufficient memory can create performance limitations later.
A balanced hardware decision is more useful than following every new technology trend.
What Does Sam Altman’s RAM Argument Mean for the Future of Technology?
Sam Altman’s comments highlight a bigger question facing the technology industry: who should control the world’s computing resources?
For decades, personal computers gave individuals direct access to processing power and memory. Users could purchase hardware, install software and decide how to use their machines.
Cloud computing introduced another model in which technology companies operate large infrastructure systems and provide computing resources as services.
Artificial intelligence is accelerating this transition. Powerful AI models require substantial infrastructure, and companies are investing billions of dollars to build the data centres needed to support them.
Altman’s argument reflects the belief that shared infrastructure can make more efficient use of limited resources. However, the opposing argument is equally important: individuals should retain the freedom to own powerful hardware and use it according to their needs.
The most practical future may combine both approaches. Cloud data centres can handle large-scale AI workloads, while personal computers continue supporting local applications, private processing, gaming and professional work.
The RAM debate is therefore not just about memory chips. It is about the balance between convenience, efficiency, affordability, privacy and individual control in an increasingly AI-driven world.
Conclusion
Sam Altman’s controversial comments about RAM have opened an important discussion about the future of personal computing and artificial intelligence. His argument is that cloud data centres can use limited memory resources more efficiently than individual computers that may leave much of their RAM unused.
However, personal computers serve purposes that cloud infrastructure cannot completely replace. Gaming, software development, privacy, offline access and local AI processing all create reasons for individuals to own powerful hardware.
The global RAM shortage also demonstrates how quickly the AI industry can influence the wider technology market. As demand for data centres increases, manufacturers, businesses and consumers will have to adapt to changing supply conditions.
Ultimately, the challenge is not simply deciding whether RAM belongs in personal computers or data centres. It is finding a balance that supports AI innovation without making everyday computing more expensive or reducing consumer choice.
FAQs
1. What did Sam Altman say about RAM?
Sam Altman argued that it is inefficient for large amounts of the world’s limited DRAM supply to remain in personal computers when that memory is used only occasionally. He believes cloud data centres can utilise shared memory resources more intensively.
2. Why did Sam Altman call RAM in personal computers inefficient?
Altman suggested that personal computers use their available RAM only around five per cent of the time. He argued that data centres could keep memory more actively utilised by sharing computing resources across multiple users.
3. Why is there a global RAM shortage in 2026?
The growing demand for AI data centres, high-bandwidth memory and advanced semiconductor components is putting pressure on manufacturing capacity. Competition between technology companies is also affecting the availability and pricing of memory products.
4. Will AI data centres increase laptop and PC prices?
Higher memory costs can increase manufacturing expenses for laptops and desktop computers. The final impact on retail prices depends on component costs, manufacturer pricing strategies, supply conditions and market demand.
5. Is local AI better than cloud AI?
Neither option is universally better. Local AI offers greater control over hardware, privacy and offline access, while cloud AI provides access to powerful computing resources without requiring expensive personal hardware.
6. Should people stop upgrading their computer RAM?
No. Users should upgrade RAM according to their actual computing requirements. Gaming, video editing, professional software and local AI applications may benefit from additional memory, while basic users may not need expensive upgrades.
