Introduction

There is no doubt that Artificial Intelligence (AI) technology is growing rapidly, and many optimists in the technology world are talking about the possibility of full-fledged AI in the coming decades. However, true sentient AI is hard to create as computers simply repeat what a human would say, making them sound sentient. This essay assumes that this barrier has been solved, and a breakthrough has led to fully sentient AI. In this case, we should be worried not about AI taking over the world, but about AI taking over our tasks, leaving most of humankind jobless.

Defenition of Sentience

First, we need to define the diverse idea of sentience in computers. Unfortunately, there is no formal definition as consciousness is an incredibly complex subject spanning across the fields of psychology to philosophy, but if the standard definition is applied to a computer, it might have an anti-climactic ending. With consciousness and no goals in life, a computer would take the logical step and understand that the best move is not to play and subsequently end. Therefore, we will proceed with the notion that the common saying of AI Sentience is better defined not as AI gaining consciousness, but more as an Artificial General Intelligence (AGI) that can replicate or exceed any intellectual task a human can and is given a goal to try and complete (Lutkevich).

Destroyer of Worlds

The concept that these advanced minds can then develop a need to destroy the world is overused and exaggerated, yet this is still grounded in some reality. Hypothetically, a highly advanced AI is given a task that can have good intentions and maximizing the task would be its goal. The AI would then calculate the most logical and best way to complete the given task and perform it. The problem comes with the ruthless efficiency that the AI would start to desire as it is optimized for such. Devoid of emotions, if something, or more worryingly someone, is in its way, the AI would prioritize the task and eliminate the obstacle to finishing it. Although we can always implement a stop button, there is a likely possibility that the robot would ignore the stop to prioritize the task (Mohan). This can be seen as the cause of the end of the world, but solutions have come forward such as Cooperative Inverse Reinforcement Learning which prioritizes not completing the task, but how a human would complete the task (Hadfield-Menell). In the end, this situation is far from reality, and a much more likely negative influence of AI is on society and especially the economy (Roose).

Societal Implications

With the progress that AI is making, there is becoming a growing threat of AGI bringing groundbreaking changes. If computers develop to the point of AGI, then the possibility of fake bots, massive editing of videos and photos, and faking of evidence becomes a reality. Employers can no longer look at a potential candidate with trust that they are real, and people chatting would constantly question if the other side is just a bot (Stewart). The most problematic, and the one that people are constantly raising awareness of, is the possibility that AGI can replace humans at jobs.

Replacing Jobs

If an AGI becomes a possibility, then hypothetically it can perfectly replicate a human, and therefore their job. Recently since this essay was made, an AI-generated artwork won first place in an art contest (Roose). If you train an AGI on the best person for the job, then the AGI would perform as if it was the best person for the job at a fraction of the time and cost a regular worker would have. This can then work every day for only the cost of electricity and be easily replicated and upgraded when needed. If this is a choice given to employers, it seems obvious that they would choose a robot instead of a human. With this in mind, employers would see the cost-effectiveness of robots as much higher and humans will slowly be phased out. While automation can replace jobs with repetitive physical tasks, a true AGI can hypothetically replace all of the jobs.

Robots replacing jobs has been a common concern throughout the ages. From textile workers going against the machinery to bank tellers worried about ATMs, economists, and experts constantly debate over the threat of innovation replacing jobs (Andrews). Fortunately for us, history has shown that despite automation, new jobs still come in, and the constant saying that people will do jobs that don’t even exist yet still hold. However, if true AGI is on the market, such a breakthrough is not simply a machine replacing a specific job, but a machine that can replicate a human and replace all jobs. If someone were to go back in history as a horse and say that innovation would only make their jobs easier, then a few years later and they’ll be proven wrong as their entire job is replaced with cars (Grey). People argue against this horse analogy, saying that horses are only there for specific reasons of transportation and nothing else. Humans on the other hand are much more diverse than horses with millions of skills that are required in jobs and more that we might not even know about. However, this argument fails if true AGI is created and humans are replicated perfectly, in that case, all our purpose would be like horses and our existence becomes obsolete.

Conclusion

Whether true AGI is coming is hard to say, but it is certainly possible. Whether you like it or not, we are simply built upon a combination of 86 billion neurons. One day, some computers would be able to replicate this and therefore become a perfect human. Whether that day is close to us is hard to say, but if it does, computers would become the new workers of society.

References

Andrews, E. (2015, August 7). Who were the Luddites? . History. Retrieved September 24, 2022, from https://www.history.com/news/who-were-the-luddites

Grey, CGP (2014, August 13). Humans Need Not Apply. YouTube. Retrieved September 2, 2022, from https://www.youtube.com/watch?v=7Pq-S557XQU.

Hadfield-Menell, D., Dragan, A., Abbeel, P., & Russell, S. (2016, November 12). Cooperative Inverse Reinforcement Learning.. arXiv.org. Retrieved September 2, 2022, from https://arxiv.org/abs/1606.03137/

Lutkevich, B. (2022, April 13). What is Artificial General Intelligence?. Tech Target. Retrieved September 2, 2022, from https://www.techtarget.com/searchenterpriseai/definition/artificial-general-intelligence-AGI

Mohan, S. (2021, October 29). Stop button paradox in agi. . Medium. Retrieved September 24, 2022, from https://medium.com/@shivamohan07/stop-button-paradox-in-agi-69c3d008ae93

Roose, K. (2022, August 24). We need to talk about how good A.I. is getting. The New York Times. Retrieved September 2, 2022, from https://www.nytimes.com/2022/08/24/technology/ai-technology-progress.html

Roose, K. (2022, September 2). An a.i.-generated picture won an art prize. artists aren't happy.. The New York Times. Retrieved September 2, 2022, from https://www.nytimes.com/2022/09/02/technology/ai-artificial-intelligence-artists.html

Stewart, J. (2019, March 1). Artificial Intelligence: The problem with perfection. Compare the Cloud. Retrieved September 2, 2022, from https://www.comparethecloud.net/articles/artificial-intelligence-perfection-problem/