What Nvidia GTC, CES, Google I/O and Dell Technologies World reveal about the next decade of computing
One of the privileges of spending a year in Silicon Valley is the ability to observe technological change as it happens. Beyond the headlines, beyond product launches and media coverage, conferences provide a unique window into how the industry thinks about the future. They reveal not only what companies are building today, but what they believe the world will look like tomorrow.
Over the past year, I had the opportunity to attend several of the technology industry’s most influential gatherings, including CES in Las Vegas, Nvidia GTC in San Jose, Google I/O in Mountain View and Dell Technologies World in Las Vegas. Each event attracts different audiences, different priorities and different perspectives. Yet despite their differences, a number of common themes emerged repeatedly. Taken together, they offer an interesting glimpse into the next decade of computing.
The first and most obvious observation is that artificial intelligence has moved beyond the experimentation phase. A year ago, many discussions still focused on what AI could potentially become. Today, the conversation has shifted toward deployment. Companies are no longer asking whether they should adopt artificial intelligence. They are asking how quickly they can integrate it into existing products, operations and workflows. At Nvidia GTC, this transition was particularly visible. The focus was no longer solely on model performance. Infrastructure, inference, enterprise deployment and industrial applications occupied center stage. AI is becoming operational.
The second theme concerns the physical world. For years, the technology industry primarily focused on digital experiences. Software consumed much of the attention. Increasingly, however, artificial intelligence is moving beyond screens and into real environments. At CES, robotics occupied a prominent place across the exhibition floor. At Nvidia GTC, physical AI became one of the most discussed concepts. Autonomous systems, industrial automation and intelligent machines appeared repeatedly throughout presentations and demonstrations. The industry seems to be converging on a new idea: the next major wave of AI may occur not inside applications but inside physical systems interacting with the real world.
This evolution naturally places greater importance on infrastructure. Throughout the conferences, conversations about AI often became conversations about GPUs, data centers, networking and energy. The software layer continues to evolve rapidly, but infrastructure is increasingly recognized as a strategic constraint. Computing capacity, power availability and cooling technologies are becoming critical variables in determining how quickly AI can scale. The industry appears to be entering a phase where infrastructure innovation becomes as important as software innovation.
Another recurring theme was the growing importance of efficiency. During the early stages of the generative AI boom, much of the focus centered on building larger models. While scale remains important, many organizations are now paying closer attention to economics. Enterprise customers want lower inference costs. Operators want greater energy efficiency. Infrastructure providers seek better resource utilization. The discussion is gradually shifting from raw capability toward sustainable scalability. This may ultimately prove to be one of the defining transitions of the AI era.
Google I/O highlighted another important trend: the integration of AI into everyday computing experiences. Artificial intelligence is becoming less visible as a standalone product and more embedded within existing tools. Rather than interacting with a separate AI application, users increasingly encounter intelligence integrated directly into search, productivity software, communication platforms and development environments. The future may not consist of a handful of AI products. It may consist of AI becoming a native layer across nearly every digital experience.
Dell Technologies World offered a complementary perspective focused on enterprise infrastructure. Many large organizations remain constrained by security requirements, regulatory obligations and operational realities that differ significantly from consumer environments. As a result, interest in private AI infrastructure, hybrid architectures and edge computing continues to grow. While hyperscale cloud providers remain dominant, enterprises increasingly seek greater flexibility regarding where workloads are deployed and how data is managed. The future of AI infrastructure appears likely to be more distributed than many initially predicted.
Perhaps the most surprising lesson across all four conferences was how frequently discussions about artificial intelligence intersected with discussions about energy. A few years ago, these topics rarely appeared together. Today, they are inseparable. Every major technology company is confronting questions related to electricity demand, infrastructure investment and sustainability. As AI systems continue to scale, the relationship between computing and energy will become increasingly important. The next decade may require as much innovation in infrastructure and energy systems as in software itself.
Taken together, these conferences suggest that we are entering a new phase of technological development. The previous era focused on connecting people through software. The emerging era appears focused on embedding intelligence throughout physical and digital systems. Artificial intelligence is moving from experimentation to deployment, from applications to infrastructure and from purely digital experiences toward interaction with the physical world.
What happens next remains uncertain. Technology forecasts are notoriously unreliable. Yet one conclusion feels increasingly difficult to ignore. The future of computing will not be defined by a single breakthrough or a single company. It will emerge from the interaction between artificial intelligence, infrastructure, robotics, energy and distributed computing systems. The most important innovations of the coming decade may not be individual products but the ecosystems that enable those products to exist.
Looking back at these conferences, the most valuable insight was not learning what technologies are being built today. It was understanding how industry leaders are thinking about the challenges of tomorrow. Beneath the excitement surrounding AI lies a deeper transformation involving infrastructure, energy, resilience and the physical foundations of computation. Those foundations may ultimately shape the future far more than any single model release ever could.
