- In your own words, describe how each of the following individuals contributed to the development of artificial intelligence. Compare your answers with the responses generated by ChatGPT.
| Alan Mathison Turing (1912-1954) | |
| In my own words
|
Turing received his doctorate from Princeton University in 1938 and worked for the Government Code and Cypher School at Britain’s codebreaking center (“Alan Mathison,” 2024). Turing devised techniques to break German ciphers, which are algorithms used to perform encryption/decryption (“Alan Mathison,” 2024). Turing was also responsible for the design of the automatic computer engine and the Manchester computer, both stored-program computers (“Alan Mathison,” 2024).
|
| Chat GPT | Alan Turing laid the foundation for artificial intelligence with his concept of the Turing Machine, a model for computation. His 1950 paper introduced the “Turing Test,” proposing that a machine could be considered intelligent if it could imitate human responses. Turing’s work established key principles in AI and computer science (OpenAI, 2024).
|
| John McCarthy (1927-2011) | |
| In my own words
|
John McCarthy was both a computer scientist as well a cognitive scientist and is considered one of the “founding fathers” of artificial intelligence (“John McCarthy,” 2024). In the summer of 1956, he attended the Dartmouth conference and together with Marvin Minsky, Nathanial Rochester, and Claude E. Shannon, he coauthored a proposal which coined the term “artificial intelligence” (“John McCarthy,” 2024). This conference was thought to be the beginning of artificial intelligence as a field of study (“John McCarthy,” 2024).
|
| Chat GPT | “John McCarthy, a pioneer in artificial intelligence (AI), coined the term “artificial intelligence” in 1956. He developed the programming language Lisp, which became fundamental to AI research. McCarthy also contributed to time-sharing systems and formalized concepts like reasoning and knowledge representation, advancing AI’s theoretical foundations” (OpenAI, 2024).
|
| Herb Simon (1916-2001) | |
| In my own words
|
Simon, in collaboration with Allen Newell, created the Logic Theory Machine, a computer program engineered to perform automated reasoning. He was considered a pioneer of artificial intelligence (“Herbert A. Simon,” 2024). He also collaborated on the General Problem Solver, a machine created to work as a universal problem solver (“Herbert A. Simon,” 2024).
|
| Chat GPT | Herb Simon was a key figure in artificial intelligence, co-developing the Logic Theory Machine and General Problem Solver, early AI programs. He explored human decision-making and problem-solving, proposing the theory of bounded rationality. Simon’s interdisciplinary work in AI, psychology, and economics significantly shaped cognitive science and AI research (OpenAI, 2024).
|
| Marvin Minsky (1927-2016) | |
| In my own words
|
Minsky developed the Society of Mind theory while at the Massachusetts Institute of Technology (MIT) Artificial Intelligence Lab (“Marvin Minsky,” 2024). This theory seeks to explain intelligence as a result of the interactions between “non-intelligent” parts (“Marvin Minsky,” 2024). Minsky visualized the brain as a machine and felt that its functions could be replicated by a computer (BBC, 2016).
|
| Chat GPT | Marvin Minsky, a founding figure in artificial intelligence, co-founded the MIT AI Lab and advanced the field with key ideas on machine learning, perception, and robotics. He developed the “Society of Mind” theory, explaining intelligence as a collection of simpler processes, influencing AI’s approach to understanding human cognition (OpenAI, 2024).
|
| Timnit Gebru (1982) | |
| In my own words
|
Timnit Gebru has worked for Apple, Microsoft, and, more recently, Google, where she co-led a team on the ethics of artificial intelligence (“Timnit Gebru,” 2024). Her employment at Google came to an abrupt end after she co-authored the paper titled “On the Dangers of Stochastic Parrots: Can Language Models Be Too Big?” This paper discussed the risks of large language models from an environmental, financial and prejudicial lens (“Timnit Gebru,” 2024).
|
| Chat GPT | Timnit Gebru is a leading AI researcher known for her work on algorithmic bias and ethical AI. She co-authored influential studies on the racial and gender biases in facial recognition systems and co-founded Black in AI, advocating for diversity and accountability in AI research and its societal impacts (OpenAI, 2024).
|
- How do “Machine (programming) languages” differ from human (natural) ones?
| In my own words
|
One of the critical differences between programming language and human language is that programming language does not change (Harris, 2018). Unlike human language, programme language cannot evolve and develop over time, and its meaning does not alter depending on the context (Harris, 2018). Take the slang terms currently trending with Generation Alpha, for example. The word “Ohio,” which indicates a US state for most adults, now indicates someone is bad, weird or cringy when placed in the context of an elementary school student. The meaning of this word evolved due to a 2016 Tumbler post (Mitchell, 2024). Human language is also both logical and emotional. If spoken, it involves body language, intonation, volume, and many other nonverbal clues (Harris, 2018). This interpretation of how something is said does not exist with programming language.
|
| Chat GPT | Machine (programming) languages and human (natural) languages differ fundamentally in purpose, structure, and interpretation. Machine languages, such as Python or C++, are designed to communicate instructions to computers with precise syntax and logic, requiring exact commands for specific tasks. They are unambiguous, with a strict set of rules that computers directly interpret or compile into binary code (OpenAI, 2024).
In contrast, human languages, like English or Spanish, are naturally evolved for communication between people. They are flexible, context-dependent, and often ambiguous, allowing for nuance, creativity, and interpretation. Humans use them to express complex ideas, emotions, and cultural concepts, unlike the rigid structure of programming languages (OpenAI, 2024).
|
- How does “machine (artificial) intelligence” differ from the human version?
| In my own words
|
In order to explain how machine intelligence differs from human intelligence, intelligence must first be defined. Chollet (2019) argues that solely measuring skill falls short of measuring intelligence. Although it requires intelligence to perform a specialized task, the ability to adapt and apply knowledge in different contexts requires a higher level of intelligence. Cholet (2019) explains how AI researchers must shift from narrow tasks to evaluating flexibility, robustness and generalization to emulate human intelligence. This can be further supported by Bloom’s Taxonomy, which illustrates a hierarchy of cognitive domains (University of Waterloo, 2024). Higher levels of thinking include analysis, synthesis and evaluation, which require the ability to think critically and problem-solve (University of Waterloo, 2024), something a machine cannot do… yet. |
| Chat GPT | Machine (artificial) intelligence and human intelligence differ significantly in their nature and capabilities. Machine intelligence, driven by algorithms and data, excels in processing vast amounts of information quickly and performing specific tasks, such as calculations or pattern recognition, often surpassing human efficiency in these areas. However, it lacks genuine understanding, emotions, and consciousness (OpenAI, 2024).
In contrast, human intelligence is characterized by complex cognitive abilities, including emotional reasoning, creativity, and moral judgment. Humans can adapt to novel situations, think abstractly, and draw on personal experiences, while AI operates within predefined parameters. Ultimately, human intelligence is more holistic and contextually aware compared to the specialized functions of AI (OpenAI, 2024). |
- How does machine learning differ from human learning?
| In my own words
|
Machine learning works on old data (Heilweil, 2020). In addition, it is often difficult to know, as a consumer, what data contributed to the “machine learning” and the degree of accuracy, bias, or prejudice within the data set (Heilweil, 2020). Although humans often refer to “old data” to learn, they have the ability to incorporate new information and analyze its relevance. Depending on the level of scrutiny and critical thinking, a human has the capacity to determine the source of data, the level of controls used, and, to a certain extent, the degree of bias impacting the data from which they are learning.
|
| Chat GPT | Machine learning and human learning differ fundamentally in processes and capabilities. Machine learning relies on algorithms that analyze vast datasets to identify patterns and make predictions, often requiring large amounts of labeled data and extensive computational power. It learns through iterations and can improve performance over time but lacks understanding or context (OpenAI, 2024).
In contrast, human learning is a complex, experiential process influenced by emotions, social interactions, and cognitive development. Humans can generalize knowledge from fewer examples, apply critical thinking, and adapt to novel situations. Moreover, human learning incorporates creativity, intuition, and the ability to reflect on experiences, enabling deeper understanding and insight (OpenAI, 2024).
|
- How do your answers to the above questions differ from what a machine could generate?
| When I started this assignment, I felt confident. I started with questions 2 through 4. I read the articles, took notes, and contemplated my opinion on the posed questions. I felt certain that the intellect of a human being exists on a higher playing field than that of a machine. The answers generated by ChatGPT for questions 2, 3, and 4 were more of a definition as opposed to my analysis and reflection on the human versus the machine. For example, each answer was scripted and presented in a similar format, beginning with a statement such as “machine learning and human learning differ” followed by “in contrast” (OpenAI, 2024). However, when I returned to question 1, the “intelligence” of the machine became evident. I could have easily spent an entire day researching the five individuals discussed in question 1. In less than 5 seconds, chat GPT was able to scour the internet looking for information and content relevant to each individual’s role in artificial intelligence. In addition to that, I could request the “machine” to summarize this content in approximately 50 words. Of course, the legitimacy and accuracy of ChatGPT’s content is undeterminable. Unlike my carefully referenced text, ChatGPT’s paragraphs provide no clear path to where the content came from, and with that, my confidence has been restored… at least for now.
|
References
Alan Turing (2024, September 24). Wikipedia. https://en.wikipedia.org/wiki/Alan_Turing
BBC News. (2016, January 26). AI pioneer Marvin Minsky dies aged 88.
Chollet, F. (2019, November 5). On the measure of intelligence
Heilweil, R. (2020, February 18). Why algorithms can be racist and sexist. A computer can make a decision faster. That doesn’t make it fair.Links to an external site. Vox.
Herbert A. Simon (2024, September 13) Wikipedia. https://en.wikipedia.org/wiki/Herbert_A._Simon
John McCarthy (computer scientist). (2024, July 10). Wikipedia. https://en.wikipedia.org/wiki/John_McCarthy_(computer_scientist)
Harris, A. (2018, October 31). Languages vs. programming languages.Links to an external site. Medium.
Mitchell, A. (2024, August 23). ‘Ohio’ is the latest Gen Alpha and Gen Z slang word – and they’re not using it in a nice way: ‘Sort of weird, absurd.’ The New York Post. https://nypost.com/2024/08/23/lifestyle/ohio-is-gen-z-and-gen-alpha-slang-for-weird-or-cringe/
OpenAI. (2024). ChatGPT (Sept 24 version) [Large language model]. https://chat.openai.com/chat
Timnit Gebru (2024, August 26). Wikipedia. https://en.wikipedia.org/wiki/Timnit_Gebru
University of Waterloo. (2024). Bloom’s taxonomy. https://uwaterloo.ca/centre-for-teaching-excellence/catalogs/tip-sheets/blooms-taxonomy#:~:text=Bloom’s%20Taxonomy%20comprises%20three%20learning,of%20the%20Taxonomy%20are%20hierarchical.