top of page
    • Jun 3, 2022
    • 5 min read

IP 2: Artificial Intelligence



1. Who were these people, and how did/does each contribute to the development of artificial intelligence? How did/does each think “intelligence” could be identified? (~50 words each)


Alan Matheson Turing was a British WWII code breaking mathematician. His work was foundational for store-program computer technology and his ideas were used as a blueprint for the first personal computers. He said that an “intelligent” mind can link “secondary, tertiary and more remote ideas” (Turing, 1950, p.18) and saw Artificial Intelligence as an “imitation”, something that can “mimic” well enough as to be indiscernible from a human.


John McCarthy was a seminal figure in the field of AI who coined the term “artificial intelligence”(Myers, 2011). He invented one of the first programming languages (LISP), which is still used today for programming AI as well as “computer time sharing”, a forerunner to Cloud-based storage/sharing. He believed in “the conjecture that every aspect of learning or any other feature of intelligence can in principle be so precisely described that a machine can be made to simulate it."


Herb Simon predicted the importance of the computer with a well-rounded understanding of far reaching consequences of technology such as climate change. Believed that “intuition is as much an expression of intelligence as is logic.”(Frantz, 2003) and that intuition was subconscious pattern recognition. He regarded the nature of human and computer information processing as being very much alike.(Frantz, 2003)


Marvin Minsky was a computer scientist and mathematician who worked with neural-network learning machines and founded MIT’s AI Lab. He believed that humans had “common sense-reasoning” and wanted to impart that to machines. He developed "frame theory" which states that when we encounter a new situation, we decide how to proceed by drawing on previous memories to guide us; a remembered framework. He believed that this theory could be extended to AI as well. (Minsky, 1975)



Timnit Gebru is a leader in AI ethics research whom was forced out of Google after whistleblowing their racist facial recognition software in a paper she wrote. Her team at google was meant to stand guard against algorithmic racism, sexism, and other bias, but afterwards she believed that oversight of Ai ethics should not be left to a corporation. About AI she said, “the biggest misconception is that it’s discussed in terms of fate—like it’s this external being that we have no control over. People need to remember that AI is something human beings create and something that we can shape in a way that doesn’t destroy society.” (Quito, 2021)




2. How do “machine (programming) languages” differ from human (natural) ones? (~100 words). For example, read Languages vs. Programming languages(Harris, 2018).


Language can be defined as a system of “spoken, manual, or written symbols that human beings use to express themselves, their identity, imagination, and emotions” (Harris, 2018). Programming languages revolve around the same principle of communication, they were created by humans as a system of “symbols and rules used to communicate a set of instructions to a machine/computer”(Harris, 2018). Morphology (contextual changes to pronunciation/meaning) is important in human languages, but not in programming languages. Programming languages are artificial; their rules were designed purposely instead of coming together organically. They either work or they don’t work, for example, there are no synonyms. Programming languages have a strict set of rules; they can’t evolve. Human language is more than just words, it includes body language, volume, intonation, etc.




3. How does “machine (artificial) intelligence” differ from the human version? (~100 words). For example, read the full article On the measure of intelligence or just the Abstract (Chollet, 2019).


The AI community generally considers intelligence to be about skills. For example, a computer besting a human at chess. However, skill is determined by prior knowledge/experience. A computer has access to unlimited practice and data prior to a game of chess. In this way, computer “intelligence” is “brittle, data-hungry”(Chollet, 2019) and struggles to make sense of new situations without human involvement. On the other hand, humans have a general intelligence; unlike AI, we are able to achieve our goals in a wide range of environments. For example, a computer program may be able to best humans in one specific video game, but a skilled human could switch to an entirely new game and still be able to apply (generalize) the skills they learned in the previous one.




4. How does “machine learning” differ from human learning? (~100 words) For example, read Why algorithms can be racist and sexist. A computer can make a decision faster. That doesn’t make it fair (Heilweil, 2020) and Artificial Intelligence Has a Problem With Gender and Racial Bias. Here’s How to Solve It (Buolamwini, 2019).


Machine learning is a kind of “training”. A program is fed massive amounts of data, then looks for patterns that it can use to make judgements or predictions. Unfortunately, the term “garbage in, garbage out” holds true; if you give the program bad data, it will replicate it. In this way, machine learning is only as good as the humans feeding it the data. For example, the program has no ability to step back and notice that the data it’s being fed is skewed against minorities, or that it’s sexist. Machines are not neutral. Humans do have the ability to shift our perspectives; we can change our responses on the fly as we learn to watch out for our own inherent biases.




5. And for your LAST challenge, a version of the Turing Test: how do YOUR answers to these questions differ from what a machine could generate? (~200 words) For this last question, think about whether your responses only reported information derived from online searches? In your responses to these questions, what transformative kinds of thinking and/or reasoning processes have you engaged in order to formulate your answers, that exceed or differ from what artificial intelligence can do? Do you think there are ANY distinguishing features that would identify your responses as having been formulated by a human, and not a machine, intelligence? What and why?


In my opinion, many of my answers are more or less what an AI could put together. The majority of my responses are drawn from the readings, just rephrased, which is something an AI could do. There are likely some human touches that an AI might struggle with, in particular the three machine learning questions, but I wouldn’t consider any of this thinking “transformative”. Perhaps an AI would have a hard time picking out what was important enough to be included in such short responses. My personal method for such assignments is to make bullet form notes as I go along, which I then use as a framework for my own responses. I’d imagine that an AI might struggle to answer a question where there was no clear answer within the given texts. To formulate a completely new idea from the clues in the literature would be higher-order thinking and require a much more sophisticated algorithm.






References


Frantz, R. (2003). Herbert Simon. Artificial intelligence as a framework for understanding intuition. Journal of Economic Psychology, 24(2), 265–277. https://doi.org/10.1016/S0167-4870(02)00207-6


Harris, A. (2018, November 1). Human languages vs. Programming languages. Medium. https://medium.com/@anaharris/human-languages-vs-programming-languages-c89410f13252


Minsky, M. (1975). Minsky’s frame system theory. Proceedings of the 1975 Workshop on Theoretical Issues in Natural Language Processing, 104–116. https://doi.org/10.3115/980190.980222


Myers, A. (2011, October 25). Stanford’s John McCarthy, seminal figure of artificial intelligence, dies at 84. Stanford University. http://news.stanford.edu/news/2011/october/john-mccarthy-obit-102511.html


Quito, A. (2021, December 15). Timnit Gebru on misconceptions about artificial intelligence. Quartz [BLOG]; Quartz Media, Inc. https://www.proquest.com/docview/2610041111/citation/DFA893D691F84755PQ/1















Комментарии

bottom of page