Human Thinking and Artificial Intelligence

Photo by Franck V. on Unsplash

Artificial Intelligence (AI) or a functioning system that can think as a human can change definition over time based on discoveries and increases in computing power or algorithms used to define intelligence. A way to define intelligence doesn't necessarily mean the system has to think like a human to be defined as intelligent. In an article titled “Artificial intelligence test: a case study of intelligent vehicles," the authors explain that a definition to describe artificial intelligence is similar intelligence or the same kind of intelligence of a human but exhibited by a machine (Li, Lin, Zheng, Wang, Liu, Cao, & Huang, 2018).

The authors explain that a computer-based AI must meet or exceed the capabilities and function of human intelligence. If we are to believe that a system has AI, they must meet or surpass human intelligence. I don't believe this tells the entire story as a toddler is still a human; however, they may not think or function as an adult would. We must consider that AI may be similar to how a human can function; however, it may not be an adult human.

In an article by Hornigold (2019), researchers have started to use how a child’s brain works as an example of creating algorithms for their artificial intelligence. The researchers believe that children's generalization capabilities are what allows them to learn quickly without continuous explanation from a parent or teacher (Hornigold, 2019). This type of research can create AI that may function similarly to a toddler.

I believe that if researchers create AI that functions like any age or capability of the human brain, we should consider it to be the standard in determining whether a system has AI. By creating algorithms that can simply pick out information in a large data set, we do not see an independent thought type of intelligence. Some of the machine learning algorithms used today sift thought significant amounts of information to find patterns. Some machine learning techniques have a computer teach itself how to walk or play a video game through a reward system. These types of systems are currently referred to as AI.  Until researchers create AI that can reason and think on its own, we are pushing the definition down the road as we create or discover more ways to measure the intelligence of a system.

References

Hornigold, T. (2019). Making algorithms more like kids: What can four-year-olds do that AI can't? Retrieved from https://singularityhub.com/2019/06/26/making-algorithms-more-like-kids-what-can-four-year-olds-do-that-ai-cant/

Li, L., Lin, Y. L., Zheng, N. N., Wang, F. Y., Liu, Y., Cao, D., ... & Huang, W. L. (2018). Artificial intelligence test: a case study of intelligent vehicles. Artificial Intelligence Review, 50(3), 441-465.