Man Vs Machine (Learning)

It seems you can't open a news site without reading an article about how "AI" and Machine Learning are going to displace humans, or the ethical implications of "AI" using human works without proper compensation. This feels like Man Vs. Machine - Part 5. This same argument was made when machines became common, and again when the steam engine gained popularity, and again when the integrated circuit was introduced, and again with the software development boom, and now again with "AI".

You may notice I use quotes around "AI". I think it's a misnomer - it's just a software program with a larger dataset and hardware that is capable of performing more diverse calculations than we are used to. At the end of the day, it is still just electrical impulses passing through circuits in a pattern designed by humans.

I wanted to check my hypothesis, that "AI" is just a rehashing of Man Vs. Machine, so I searched through old news paper articles. There's a pretty cool site called Library of Congress - Chronicling America that has scans of newspapers from 1777 to 1963. It currently has over 20 million articles, and doing a quick search with the keywords "machine replace human" shows results dating back to 1885.

Interesting enough, the earliest article is from 1885 and has a section about machinery breaking at a lighthouse and how a family banded together to replace it. They gave up sleep and manually operated the machine for 5 weeks to ensure that it stayed on its regular 90 second cycle to keep ships safe at sea. This article is a sneak preview into how machines replacing humans isn't flawless, and we need to have backup measures.

Coincidentally, this seems to be a strong argument for Right-to-Repair legislation is important. What do we do when the machines break?

As early as 1920 we can see advertisements for farm equipment explicitly saying that they're great for replacing human, horse, and mule power.

We also see articles referencing the "human machine", referring to people doing manual labor. Some articles say human are better, because we self-repair and don't break down like machines do; other articles say humans are worse because they can't work as long or as hard.

I remember being in grade-school and seeing cartoons from the 1950's about American Folk Legends Paul Bunyan, a lumberjack, and John Henry, a steel-driving man, competing against the steam engine - and both inevitably losing. However, they are still fondly remembered. They didn't become useless when the steam engine came, their purpose simply changed.

These articles that are hundreds of years old show that nothing is really new - especially not the true impacts of these advancements. An article from 1926 talks about the impacts of mining equipment: the owners saw a 2x improvement in output, and the laborers had their wages cut 10%. The owners of the business (and machinery) see more profit, and the workers see less profit. This trend is common in business from all aspects, not just technological improvements. To me, this is the real issue: who reaps the benefits from these improvements, and can people still make a good living?

An article from 1932 attempts to put this in a positive light. It is titled, "Five-day working week inevitable". Think about that, less than 100 years ago there was talk about us only working 5 days a week, and today in 2022 there are studies being done that say a four-day working week is really all we need. This article summarizes what machines are actually meant to do:

The trend of mechanical invention and of machine replacement is to relieve the human hands of their most intricate and skillful efforts in the production process.

and goes on to identify how society has responded and the ideal solution:

The attitude of society toward the machine since its first installation in industry has been to more and more control operation, and to restrict and limit its competition with human effort.

It's fine that technology replace jobs - it's how it is done that matters.

You may also notice that I'm not spending much time talking about "AI" or Machine Learning specifically in this power. That's because it doesn't matter much to me, they're both just technologies. They will eventually gain mass acceptance, fade into obviousness, and then a new technology will become the hot topic for discussion.

Change is constant. Technology will replace people's jobs, but not people. The only sure way to become obsolete is to spend your time fighting to keep your old position, instead of improving and adapting to new ways.