What impact will artificial intelligence have on human society?
The first consequence is that mankind will become smarter. If AI-powered simulation games are available and AI-powered search engines are crawling for existing information, it is easier for humans to get an accurate picture of the world. The result will be that the number of scientists will increase. That means it will become normal to be familiar with the latest research questions and writing books about it. A second effect will be that the price of goods and services will become lower. In the computing industry, this is known as the Moore's law, but it can be transferred to the car industry too. And a third (negative) effect will be that the paranoid level of individuals will increase. In a world, dominated by robots, knowledge and computers some kind of counter-movement will take place in which people starting to construct religion-driven explanations because they are overwhelmed by the reality.
This article on Medium can give you a basic understanding of the history of AI. Basically, Alan Turing, the father of AI, first coined the concept that an aspect of intelligence could be so precisely described that a machine could be made to simulate it.
My sense is that AI begins with automation. Originally I believed water clocks were the first embodied algorithms, but now I think the first simple traps and snares: Are simple animal snares and traps a form of automation? Of computation?
(Mechanism, in and of itself, is not understood to be intelligence, but when mechanism is selected for fitness, it demonstrates it.)
In terms of the formalization of the field, that began with Game Theory, Turing, Shannon, et al., and the 1956 Dartmouth conference.
My underlying point is that automation seems to have been around as long as homo sapiens sapiens, and the impact has been profound in all endeavors related to utility.
If water clocks are the first primitive AIs, then measuring time was the impact, which enhanced our ability to manage various systems. If simple traps and snares, the impact was more food and greater fitness for the human species in our environment.
If AI requires making decisions (choosing between alternatives,) not just modifying or responding to input (as in a simple reflex agent,) that probably arose in the first simple agricultural equipment for sorting produce, separating wheat from chaff, where the intelligence is built in to the artifact.
Here it would have resulted in significant savings of time and freed up humans labor for other endeavors, allowing civilization to flourish. (Civilization flourishing freed up time for humans to think up new artifacts that increase utility, and methods for calculation and storing data, abaci and clay tablets initially, then paper, then magnetic tape, etc., and way better calculators with logic functions built in.)
If we're talking about contemporary AI, in the sense of Machine Learning and Genetic Algorithms, where processing power and memory only increase, we can expect the effects of AI to be more profound than we can imagine, and have numerous impacts we will be unable to predict.
If the history of technology gives any indication, the impacts are likely to be a mix of positive and negative, beneficial and harmful.