The thing is, we've only really been seeing this boom since the 18th century. And for most of the world, not really until the 20th century. The solution, as you said, for most of this period was to have people consume more products. But with the rise of electronics, that trend doesn't appear to be continuing. We already live in a world where not everyone needs to work: in fact, we specifically push people out of working once they reach retirement age. That was one of the points of Social Security and pensions: to stop older workers from preventing new workers from taking their place.
Also note, our real unemployment isn't 7.9%. In fact, if we look at the number of non-farm jobs from the last census in 2010, we had 112 million jobs. If we assume about 3% of employment is farm jobs, then we have ~115 million total jobs. According to the census, there are 313 million people in the US, and 76.3% of them are adults. Therefore, we have 238 million adults, and only ~115 million jobs. This means our real unemployment rate is about 52% (given the roughness of my figures, I'd stick it at a range between 45 and 60 percent). Which means, not counting children, our country is already fully capable of supporting half the population not working. This wasn't the case in 19th century America (where yes, even women usually worked in some capacity), meaning we probably already are in a society where most people don't have to work, we just don't notice it because most of those who don't work are not labeled "unemployed."
Edit: Basically, I'm questioning your premise that employment rates have, in fact, remained relatively constant throughout the last few centuries.
You are questioning whether the current Labor Force Participation rate and whether it has gone down over the years. The peak in recent history appears to have been around 2000 at 67.3%. Currently, it's at 63.6%.
The furthest back I could find is 1948 for 58.6 at the start of that year. Last time we were in the 63% range was in 1980s. Guessing from 19th century, it has gone up.
The key here is to understand what influences this rate. Many factors go into what percentage of population chooses to seek employment. For example, if my wife and I are rich enough, when we have kids, we may want to have one or even both of us take a few years off to raise them. This is a profound luxury in the current US society. Some countries have extended maternity leave that lasts for 3 years and includes some form of pay or other welfare payment. Such structures would easily drive the participation rate down.
Conversely, public daycares and other child-friendly services promote higher participation rates by freeing both adults to work.
These examples are meant to show that the "machines are taking our jobs" discussion to be a gross oversimplification of a much more complex system. So far, history tells us that we will adapt. That still leaves a possibility of a black swan event... In either case, I have faith in humanity :-)
Also note, our real unemployment isn't 7.9%. In fact, if we look at the number of non-farm jobs from the last census in 2010, we had 112 million jobs. If we assume about 3% of employment is farm jobs, then we have ~115 million total jobs. According to the census, there are 313 million people in the US, and 76.3% of them are adults. Therefore, we have 238 million adults, and only ~115 million jobs. This means our real unemployment rate is about 52% (given the roughness of my figures, I'd stick it at a range between 45 and 60 percent). Which means, not counting children, our country is already fully capable of supporting half the population not working. This wasn't the case in 19th century America (where yes, even women usually worked in some capacity), meaning we probably already are in a society where most people don't have to work, we just don't notice it because most of those who don't work are not labeled "unemployed."
Edit: Basically, I'm questioning your premise that employment rates have, in fact, remained relatively constant throughout the last few centuries.