That is actually the definition of 'DARPA - Hard'. Maybe DARPA ought to get in on this Mars colonization business...maybe float a DARPA Grand Challenge.

Smart phones capabilities may be nearing a 'local maximum', but the real revolution is waiting to happen in the invisible half of the cellular ecosystem - the cellular network. Smart phone ownership is already very high in the developing world, and 'access' to a smart phone is almost always there. But cellular data rates are still pretty much un-affordable to most people in the developing world. Data connectivity is also still quite patchy. Sure you can purchase a dirt cheap pre-paid data pack and get 'connected'. But what good is it if you get GPRS speeds on a '3G' network? Or if your pack runs out in 3 days, when you try to do something meaningful with it (say participate in a distance learning class, while still working on your family farm, 10 kilometers away from the nearest school)?

They should repeat this experiment at Def Con 2016.

Couldn't agree more. You really need four things to code effectively: 1. Logical reasoning skills, 2. procedural thinking skills, 3. A good tool that is expressive enough for what you want to accomplish, and doesn't 'get in your way', and 4. A desire to code! It doesn't matter whether you are writing each line of code, or doing drag-n-drop. All those are ultimately visual representations of the underlying thought process.

5G is supposed to codify a *set* of objectives that networks need to satisfy. These include: .

• Mindnumbingly high throughput when the user is stationary or moving slowly.
• Very high throughput even when the user is moving very fast (e.g.high speed train travel upto 300 kmph).
• Very low latency connections (1-10 ms)
• Support for massive city-wide or region-wide deployments of Internet of Things over low power cellular connections
• Very high spectral efficiency leading to reliable high throughput connections in densely populated neighboords

These are some of the important objectives. Speed is just one of the criteria for qualification. In that sense, the test described in the article seems to be more like 4.1 G in the same way plain LTE (i.e. not LTE Advanced) is 3.5G.