Scientists at Virginia Tech argue that the wireless revolution—think 6G and beyond—demands artificial intelligence (AI) capable of human-like reasoning. The critical ingredient? Common sense: the ability to comprehend the world and adapt to new scenarios. This perspective comes from an IEEE Journal article authored by Professor Walid Saad, graduate student Omar Hashash, and postdoctoral fellow Christo Thomas. AI today consists of programs that learn and solve problems, but its capabilities are narrow. It excels at spotting patterns in data—identifying a dog in a photo, for instance—but struggles to grasp context or predict outcomes, like what happens if that dog chases a ball. Common sense would enable AI to anticipate such events, much like humans do intuitively. Emerging technologies, such as the metaverse—a virtual world mirroring reality—rely on networks that seamlessly blend the physical and digital. This requires AI with human-like cognition. To achieve this, the researchers propose leveraging “digital twins,” virtual replicas of real-world objects or systems. Imagine donning a VR headset and “walking” through a city while the network adapts in real time, as if you were physically present. That’s the vision. Professor Saad predicts that artificial general intelligence (AGI)—AI with human-level reasoning—remains 10 to 15 years away. “We’re looking at a decade or more before wireless networks integrate general-purpose AI capable of thinking, planning, and imagining,” he explains. “The metaverse and digital twins will accelerate this by modeling the world and helping AI develop common sense.” So, what’s holding us back? Current AI is tethered to statistical analysis, faltering when faced with unpredictability or tasks requiring creative problem-solving. For instance, it can’t figure out how to connect a phone to a weak-signal network without explicit instructions. The scientists call for a paradigm shift: from networks aided by AI to networks that think independently. “We need wireless systems that not only grasp AI’s underlying principles but also understand the world—its objects and their interactions,” says co-author Christo Thomas. “This demands integrating mathematics, reasoning frameworks, and insights from cognitive science.” A digital twin of a wireless network could bridge this gap, enabling AI to perceive and interpret reality as humans do. It’s the foundation for a network that doesn’t just function but anticipates and plans ahead—essentially, the “missing link” for intelligent systems. The researchers envision a future where your router optimizes signal strength on its own or the network predicts where you’ll need connectivity next. This isn’t just about faster data; it’s about technology that understands us. In this scenario, the metaverse evolves beyond a gaming platform into a training ground for AI, fostering the leap to true intelligence. The post Virginia Tech: AI Must Think Like Us for Next-Gen Networks appeared first on Anomalien.com.