A more interesting narrative is that our tools have become too complicated for most people to understand and use efficiently. The cost per km of the first Lond

A more interesting narrative is that our tools have become too complicated for most people to understand and use efficiently. The cost per km of the first London Underground lines, which were dug largely by hand from the road downwards and built with a lot of manual labour was about 1/10th of the cost of building a similar tunnel in the 21st century with automated machinary and far fewer "expensive" people. How do you dig by hand? Simple, get a load of people with no training and give them a spade and get on with it. How do you do that with a $100M tunnel boring machine? Send people on expensive training courses, pay a load of cash in specialist repairs and maintenance and when it breaks down or you get any other kind of problem off the beaten track, just stop work until you can sort it out or pay someone else even more money to get you out of your hole. Same with software. How easy is it to keep bank records in a physical ledger? Really easy and very quick. Introduce some software and you now have a multi-million dollar complicated recording machine with loads of semi-skilled operatives using it, breaking it (accidentally), and only very slow and expensive teams can investigate and fix. We have accepted that sometimes we need to ask 10 times to get an address changed because "computer says no". Of course, technology brings you benefits that didn't exist before like centralising data, like making some hard physical processes easier to automate and to backup data that could otherwise be lost in flood or fire but are these worth the complexity and cost? In many cases that's a hard "no"!