Automation

Automation is the use of computers to control industrial machinery, replacing human operators. It is a step beyond mechanization, where human operators are provided with machinery to help them in their jobs. The most tangible part of automation can be said to be industrial robotics.

By the middle of the 20th century, automation had existed for many years on a small scale, using mechanical devices to automate the production of simply shaped items. However the concept only became truly practical with the addition of the computer, whose flexibility allowed it to drive almost any sort of task. Computers with the required combination of power, price, and size first started to appear in the 1960s, and since then have taken over the vast majority of simple assembly line tasks (some food production/inspection being a notable exception).

Social issues of automation

Automation raises several important social issues. Among them is automation's impact on employment/unemployment.

Some argue automation leads to fuller employment. One author made that case here: When automation was first introduced, it caused widespread fear. It was thought that the displacement of human workers by computerized systems would lead to unemployment (this also happened with mechanization, centuries earlier). In fact the opposite was true, the freeing up of the labor force allowed more people to enter information jobs, which are typically higher paying. One odd side effect of this shift is that "unskilled labor" now pays very well in most industrialized nations, because fewer people are available to fill such jobs.

Some argue the reverse, at least in the long term. First, automation has only just begun and short-term conditions might partially obscure its long-term impact. (E.g., during the US IT boom of the 1990's, fewer Americans noticed the "giant sucking sound" of manufacturing jobs leaving America.) Second, marketing new automation solutions (currently) requires significant human intellectual capital, especially at the outset of the effort. But, after the railroads were built, did railroad companies need the same number of railroad workers? No. Third, although those same railroad workers were able to find other jobs at the time, does "past performance guarantee future results"? In the next 50 years, we may well develop AI agents with IQ's of around 130 that we can "duplicate" for the cost of, say, two yearly human salaries. At that point, humans will have only incidental real economic value to large enterprises -- regardless of how hard they study, how smart they are, and how hard they work. The richest investors/executives will hire/retain fewer "knowledge workers" (at least, of the human variety). In sum, at least one prognosticator expects automation to a) turn the relatively new "jobless recovery" phenomenon into a recurring event and b) accelerate the economic stratification in the US (started in the 1980's through a flattening of tax rates).