Most of the planning for our infrastructure is around preparing and managing peaks. Electricity generation capacity is usually based on the load of all the air conditioners running on the hottest days. Six lane highways are built to move as many cars as possible during rush hour. Sewers need to handle sever storm waters.
The result is that most of the time, infrastructure is under utilized, sometimes to a gross degree. The problem is exacerbated with a centralized infrastructure, where the peaks can’t be smoothed. The six lanes of traffic doesn’t help if all the one lane exits are backed up for miles.
To handle this, systems that can use this, can switch to distributed generation. If your neighbour has energy generation capacity when you need it, the transportation is reduced as well as the inefficiency and losses. System capacity can also be added at a much more discrete level.
With IoT and information processing, we have even more processing capability built into unused devices. Amazon has released tools for a few years to push processing to the edge as much as possible to decrease response times. What might be the next stage is to use IoT devices’ capacity for processing rules and even preliminary speech recognition.
Our Echos and Homes have lots of unused processing capacity and most of the millions of devices on the market are not being used concurrently. What if processing could be distributed to the other devices that are close by?
The future that we might see come quickly is one where the Internet is even more ubiquitous. All devices not only connect to it but also process it. Maybe even that’s the real communism — but only among machine: each works to its own ability.