5G continues to generate headlines. All the talk about 5G radios is interesting, but those radios are only part of the 5G story. As I dig deeper, the story becomes stranger and stranger, with the radios distracting us from the issues of 5G networking protocols and policies. I'm concerned about the risks of accepting the idea that we need a 1970s style telecommunications network. It's the triumph of marketecture over architecture. Why isn't that story being covered? There is a risk in treating the Internet as just another telecommunications service (relegated to the slow lane). It's just the opposite—or should be. A phone call is just an app and not a network service. What happened to all we've learned about best-efforts packet connectivity? Why is our policy at odds with reality? The consequence is to limit our ability to communicate and innovate. Another risk is expertise creep. I respect the expertise of radio engineers. But that doesn't mean that they are experts in the software and business protocols for connected devices and applications. Remember that telecom engineers told us we needed a special network for voice until VoIP happened. Today we're again being told that we need a special network for applications such as video and connected devices even though we're doing just fine without one. More to the point, we're doing just fine because we can innovate outside of the network, and that's a problem for the legacy business model. Requiring a SIM cheap creates unnecessary dependencies and opportunities for failure. I could go on, but there is so much weirdness that I wrote a whole column asking why the IEEE has fixated on 5G as the one future. For the deep dive into 5G https://rmf.vc/IEEE5GPast.