Will AI lead to the destruction of society?
This article participated in the Immerse Essay Competition and was awarded a partial scholarship to attend a summer program at the University of Cambridge.
Mankind is the most remarkable being among all species because he possesses a characteristic that others lack: complex thought. This superior form of intelligence has catapulted humans to the forefront of society allowing them to mold earth into whatever form deemed most propitious. However, this supremacy could be undone by humanity's own creation: Artificial Intelligence (AI). This technological advance, while fascinating at first glance, has the potential to subjugate society and change its entire socioeconomic structure. Eliezer Yudkowsky, a researcher in the field of artificial intelligence, introduced the concept of DWIM, which stands for Do-What-I-Mean and states that programs do what you tell them to do, not what it is implied to do (Yudkowsky, 2008, p. 13). AI will do exactly what programmers ask it to do to achieve the specified goal, even if the course of action is not what humans want. For this reason, a program's instructions must be as clear as possible, otherwise, it could lead to disastrous consequences. For example, Tesla issued a statement when one of its customers died because its Autopilot crashed into the side of a truck (Tesla, 2016, n.p). Since this AI was developed to recognize vehicles on a highway, and since it is very rare for trucks to be on their sides on a highway, it may not have recognized it as a vehicle. This shows that AI performs tasks oblivious to the damage it may cause. With that being said, humans will continue to improve technology and reach a point where machines will surpass them intellectually; therefore, today's workforce will be drastically different from those in a few years. Anthony Goldbloom, CEO of Kaggle, a competition for data scientists, spoke about the future of jobs. He pointed out that there are programs that can evaluate student work with the accuracy of a high school teacher (TED, 2016, 1:43). This suggests that jobs that are too repetitive and require little thinking can easily be replaced by machines. Albeit, Goldbloom stated that humans can handle novel situations, while machines cannot. Such a statement, however, underestimates the power of AI. Softwares like Jasper.ai can create blog posts, captions, and even marketing copy. Therefore, it is very likely that AI will be able to do any task - even more complex ones - in the coming years. If it does not gain this ability, which is quite unlikely, it will wipe out employment opportunities for those who do not have the financial means to pursue studies for more complex and higher-value tasks, leading to a socioeconomic wealth gap and an increase in global hunger. The reason AI is frightening is because the world is still uncertain about its full capabilities. The increasing speed at which technology is developing will certainly lead to a creation that will be difficult to control. When something like this happens, society will have to ensure that the machine's values and goals are perfectly aligned with those of humanity.
Goldbloom, A. (2016). " The jobs we'll lose to machines -- and the ones we won't". TED.
The Tesla Team. (2016). "A Tragic Loss". Tesla. https://www.tesla.com/blog/tragic-loss
Yudkowsky, E. (2008). Artificial Intelligence as a Positive and Negative Factor in Global Risk.