The Truth About 5G: What Cellular Networks Really Deliver

01 The Promise and the Reality

When 5G was first pitched to consumers, the claims were astronomical: download a full movie in seconds, enable self-driving cars through instant communication, replace your home internet with a wireless connection. The marketing implied that 5G was a single technology delivering a single massive upgrade. The reality is more complicated. 5G is not one thing — it is a collection of technologies operating across different frequency bands, each with dramatically different capabilities, and the experience depends almost entirely on which band your phone happens to be connected to at any given moment.

The gap between promise and delivery was large enough to spawn conspiracy theories. The idea that 5G caused COVID-19, or that its towers were surveillance devices, or that the millimeter-wave signals could cook human tissue — none of these are true, but they flourished in an environment where the technology was oversold and under-explained. The truth is that 5G is a genuine and significant improvement over 4G, but it is not magic. Understanding what it actually does requires understanding the radio spectrum it uses and the infrastructure it demands.

02 How 5G Actually Works

At its core, 5G is a set of radio communication standards defined by the 3rd Generation Partnership Project (3GPP), the same standards body that defined 3G and 4G LTE. The fundamental advance is not a single breakthrough but a combination of improvements: wider channel bandwidths, more efficient modulation schemes, massive MIMO (multiple-input multiple-output) antenna arrays, and beamforming — a technology that directs radio signals precisely at individual devices rather than broadcasting in all directions. Together, these technologies allow a 5G base station to serve more devices simultaneously, at higher data rates, with lower latency than a 4G LTE tower.

The latency improvement is real and significant. 4G LTE typically achieves round-trip latency of 30 to 50 milliseconds. 5G, in optimal conditions, can achieve 1 to 10 milliseconds. For most consumer applications — streaming video, web browsing, social media — this difference is imperceptible. But for industrial applications — remote surgery, autonomous vehicle coordination, industrial automation — the reduction from 50 to 5 milliseconds can be the difference between a system that works and one that does not. This is where 5G's value proposition is strongest, and it has almost nothing to do with the consumer smartphone experience that dominates marketing.

03 Frequency Bands and Spectrum

The single most important thing to understand about 5G is that it operates across three very different frequency ranges, and the band determines the experience. Low-band 5G, operating below 1 gigahertz, provides coverage similar to 4G LTE — signals travel several kilometers and penetrate buildings reasonably well — but offers only a modest speed improvement. This is the 5G that most people actually have on their phones most of the time, and it is why many consumers feel underwhelmed: their phone says 5G but the experience is barely distinguishable from LTE.

Mid-band 5G, operating between 1 and 6 gigahertz, is the sweet spot. It offers a genuine speed improvement over 4G — typically 3 to 5 times faster — with reasonable coverage of about 2.5 kilometers per tower. This is the band that most carriers are deploying aggressively, and it is where the real consumer benefit of 5G lives. Mid-band also supports the channel widths needed for meaningful capacity improvements, allowing a tower to serve more simultaneous users without degradation.

Then there is millimeter-wave (mmWave), operating at 24 to 40 gigahertz. This is the 5G of marketing demos — gigabit download speeds, ultra-low latency, the technology that can theoretically download a movie in seconds. The catch is that mmWave signals have a range of roughly 300 meters, are blocked by walls, trees, rain, and even human bodies, and require a density of small cell installations that no carrier has been willing to fund at scale. mmWave 5G exists in pockets — sports stadiums, downtown cores, specific neighborhoods — but it is not, and likely will never be, a ubiquitous cellular technology.

04 Small Cells and Infrastructure

The infrastructure requirements for full 5G deployment are staggering, particularly for mmWave. A traditional 4G macro cell tower covers a radius of several kilometers. A mmWave small cell covers roughly a city block. To blanket a metropolitan area with mmWave would require thousands of small cell installations on utility poles, streetlights, and building facades — each requiring power, backhaul fiber, permits, and maintenance. The cost is enormous, and carriers have largely concluded that the return on investment does not justify a full mmWave buildout outside of high-traffic areas.

Instead, most 5G deployments rely on mid-band spectrum and existing tower infrastructure, upgraded with massive MIMO antennas and beamforming capability. This approach delivers most of the practical benefit of 5G — faster downloads, more capacity, lower latency — at a fraction of the infrastructure cost. The result is that 5G, as most people experience it, is essentially a better 4G: the same towers, roughly the same coverage, but with better spectral efficiency and more capacity per cell. This is a genuine improvement, but it is not the revolution that was promised.

05 Real-World Performance vs Marketing

The gap between marketing claims and measured performance is where the Business Insider investigation that informs this article lands its sharpest criticism. Carriers advertise 5G speeds of "up to 1 gigabit per second," but OpenSignal's independent testing consistently shows median real-world 5G download speeds in the United States ranging from 75 to 200 megabits per second, depending on carrier and location. That is faster than 4G LTE, which typically delivers 25 to 50 megabits per second, but it is nowhere near the gigabit promises. The "up to" qualifier does enormous work in these advertisements.

Part of the problem is that carriers label all three bands as "5G" on the phone's status indicator, even when the phone is connected to low-band 5G that is barely faster than LTE. A consumer seeing the 5G icon expects a dramatically different experience and instead gets something marginally better. This labeling ambiguity has been the source of considerable consumer frustration and regulatory scrutiny, but the carriers have resisted requests to distinguish between 5G bands in the user interface, arguing that the distinction would be confusing — or, more cynically, that it would reveal how little actual 5G benefit most users are receiving.

06 5G and the AI Revolution

The connection between 5G and artificial intelligence is less direct than the connection between, say, 5G and autonomous vehicles, but it is growing. Edge AI — running machine learning inference on devices rather than in cloud data centers — depends on low-latency connectivity to function effectively. A factory robot equipped with cameras and sensors needs to send data to a local edge server for real-time inference, and 5G's low latency makes this possible in ways that 4G LTE's 30-50 millisecond round-trip time does not. The industrial IoT use case — where 5G replaces wired Ethernet in factories, ports, and warehouses — is where the technology delivers on its promises most clearly.

For consumer AI applications, the impact is more nuanced. Large language models run on cloud servers regardless of the connection speed; the bottleneck is compute, not network. But as AI models move toward real-time multimodal interaction — processing video, audio, and text simultaneously — the bandwidth and latency improvements of 5G become relevant. A real-time AI assistant that analyzes a live video feed needs a connection that can carry high-bandwidth sensor data with minimal delay. 5G mid-band, with its combination of reasonable latency and higher throughput, is well-suited for this class of application, though the real-world deployment is still in its early stages.

07 What 6G Might Bring

Even as 5G continues its slow rollout, the industry is already defining 6G. The 3GPP and ITU have begun early standardization work, with a target commercial deployment around 2030. The proposals include terahertz-frequency communication (100 gigahertz to 1 terahertz), which would offer peak data rates of 100 gigabits per second — a hundred times faster than current 5G. The challenges are even more daunting than those that faced 5G: terahertz signals have a range measured in meters, not kilometers, and are blocked by virtually everything, including humidity in the air.

More practically, 6G is expected to focus on integrating AI natively into the network stack — using machine learning for dynamic spectrum management, beam optimization, and traffic routing. The vision is a network that does not just carry AI traffic but is itself an AI system, continuously optimizing its own performance. Whether this vision materializes on schedule is an open question — 5G taught the industry that technology transitions take longer and deliver less than the marketing implies. But the direction is clear: the future of cellular networks is not just faster data, but smarter infrastructure, and the line between the network and the applications it carries will continue to blur.