It is imperative to challenge the definition of AI as stated at the beginning of this paper. This definition implies that all computer systems are, effectively, AI systems because they are simply getting more and more intelligent. However, they are getting more intelligent because humans are, and have up until the present, been creating them that way. This, therefore, is an incomplete definition of AI. Computer intelligence should not be measured by how much a device knows, but by its ability to gain knowledge. To use the paper's example, a calculator can perform calculations much quicker than the human brain. However, this calculator does not draw from memory to recognize numbers and calculations as humans do. Each time the calculator undergoes a calculation, it treats it as an entirely new operation, much as a child learning basic math might do. The reason we consider these calculators to be "dinosaurs" in the face of modern computers is because modern computing systems, unlike calculators, are able to teach themselves. Therefore, a system cannot be considered AI until it has the ability to teach itself. Without this ability, its intellectual capacity is capped by its programmers. 

Additionally, this paper frequently mentions the social ramifications of artificial intelligence, claiming that AI will lead to job loss. Before this problem can be solved, it is important to address the basis of these claims. Why do people fear AI taking their jobs? It is the concept of adaptability. Previously, machines were tailor-made to their jobs. Even computers have limited intelligence, and therefore perform their functions well and do not intrude too much into the human lifestyle. However, as AI machines become more sophisticated, they are able to adapt to a changing environment and learn how to accomplish tasks they were not initially told to undertake. 

There are a few different solutions to this problem. Many AI experts such as Mark Zuckerberg and Elon Musk claim that we will need to establish greater social rights such as a universal basic income (UBI). However, though this suggestion may sound promising, it will be notoriously difficult to implement and maintain, and will no doubt fail at some point in the future. The issues that a UBI aims to mitigate are a tangible threat: the loss of low-level jobs that are currently held by humans. A UBI will not abate the loss of these jobs. Instead, it will make people overly dependent on it, and its costs will inevitably swell to become too large to sustain. Therefore, it is imperative that instead of leaping towards this seemingly elegant solution, we overcome this hurdle by bolstering our own human education. 

The best way to overcome the issue of computer adaptability is to bolster human adaptability through education. It is shocking that we teach AI machines how to learn and not our own human children. Currently, education (especially higher education) tends to be tailored to one specific field or job. This leads to people becoming more and more compartmentalized within a certain field, and even more susceptible to job loss by AI systems. Therefore, instead of adopting the UBI model that this paper alludes to, we should be focusing on adding the concept of adaptability to our education systems. This simply translates to teaching students how to learn. This will and should shatter the long-held paradigm that education happens solely in the classroom. The more we grow accustomed to constantly educating ourselves to an ever-changing world, the less likely we are to be marginalized by it.