Table 1. A side-by-side comparison of my responses to a series of questions vs the responses of Chat GPT (OpenAI, 2023).
| My Response |
Chat GPT’s Response |
| 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 Turing |
|
| Alan Turing was a British mathematician who paved the way for computer sciences and the development of artificial intelligence (Biography.com, 2020). He proposed a machine that could compute anything and created the design that led to the first personal computers. He described an Imitation Game that would test machine intelligence by observing it in conversation with a human (Turing, 1950).
|
Alan Turing was an English mathematician and computer scientist who helped crack Nazi codes during WWII and is considered the father of computer science. He proposed the concept of a machine, now called a Turing machine, that could perform any mathematical calculation, laying the foundation for modern computers. Turing believed that if a machine could successfully mimic human conversation, it could be considered intelligent. This became known as the Turing test. |
|
John McCarthy |
|
| John McCarthy, a professor of computer science, was a researcher of artificial intelligence and the creator of Lisp programming language. He defines intelligence as “the computational part of the ability to achieve goals in the world” but believes it is not yet possible to define machine intelligence without relating it to human intelligence (McCarthy, 2007).
|
John McCarthy was an American computer scientist and mathematician known as one of the founders of artificial intelligence. He helped develop the first AI programming language, LISP, and organized the Dartmouth Conference in 1956, considered the birth of AI as a field. McCarthy believed intelligence could be defined as the ability to solve problems and make decisions, and that machines could be made to demonstrate this through programming and algorithms. |
|
Herb Simon |
|
| Herb Simon was a researcher of management science who focused on decision making. His work on psychology, economics, computer science, and math made him one of the founding fathers of artificial intelligence (Donovan, n.d.). His programming was informed by a computational theory of human intelligence where large amounts of data are collected to understand the background of a problem (Lee, 2013).
|
Herbert Simon was an American economist and psychologist who was awarded the Nobel Prize in Economics in 1978. He made significant contributions to the study of decision-making and artificial intelligence. He believed intelligence could be identified by a system’s ability to process information efficiently and make decisions that achieve its goals. He also introduced the concept of “bounded rationality”, suggesting that decision-making is limited by the information available and the cognitive abilities of the decision-maker. |
|
Marvin Minsky |
|
| Marvin Minsky, cofounder of the Massachusetts Institute of Technology’s Artificial Intelligence Lab, built one of the first neural network machines and strove to design machines that were capable of human common-sense (BBC, 2016). His theory of intelligence was based on the idea that many small non-intelligent agents can work together in a society to produce something intelligent (Minsky, 1986).
|
Marvin Minsky was an American cognitive scientist and computer scientist who is considered a pioneer in the field of artificial intelligence. He co-founded the MIT Artificial Intelligence Laboratory and contributed to the development of early AI systems. Minsky believed intelligence could be identified by a machine’s ability to use symbols and rules to represent and manipulate knowledge, leading to “common-sense reasoning”. He also developed the theory of “frames”, which is used to represent knowledge in AI systems. |
|
Timnit Gebru |
|
| Timnit Gebru, a computer scientist and artificial intelligence ethics researcher, has analyzed representation of women, people of colour, and less developed countries in AI-produced output (Hao, 2020). The work that caused her ousting from Google outlined the risks of large language models. She believes that intelligent models of AI would be built with community contexts in mind (Walsh, 2022).
|
Timnit Gebru is an Ethiopian-American AI researcher and computer scientist who has made significant contributions to the field of artificial intelligence ethics and bias. She co-led Google’s ethical AI team before leaving the company in 2021. She advocates for responsible AI practices and has published research on the unintended consequences of biased data and algorithmic decisions. Gebru believes intelligence can be identified by a machine’s ability to learn and understand complex patterns in data and make decisions that align with human values and ethics. |
| 2. How do “machine (programming) languages” differ from human (natural) ones? (~100 words) | |
| While both machine and human languages are forms of communication with semantics and syntax, they are different. Harris (2018) outlines three major differences: 1) Human languages can change depending on context, while machine languages have defined rules and do not change; 2) Natural languages include nonverbal communication that can impart emotion and nuances that are not possible to convey in machine language; and 3) in human language, errors, accents, slang and other forms of improvisation do not necessarily affect the meaning, while in machine language, these things may cause errors and render the communication ineffective. | Machine languages are designed for computers to understand and execute, whereas human languages are for communication between people. Machine languages are based on strict syntax and grammar rules, whereas human languages can be more flexible and context-dependent. Machine languages consist of a limited set of commands and symbols, whereas human languages have a large vocabulary and the ability to form new words. Machine languages are designed for efficiency and speed of processing, whereas human languages have evolved for nuanced communication and expression of emotions. Machine languages can only understand specific tasks, whereas human languages can understand and convey abstract concepts. |
| 3. How does “machine (artificial) intelligence” differ from the human version? (~100 words) | |
| Machine intelligence is often measured based solely on the efficiency and speed with which the AI can perform a skill, whereas human intelligence is informed by prior experience (Chollet, 2019). Traditionally, intelligence in AI was determined by success in a specific task, but human intelligence can be applied generally, over multiple different tasks. According to Chollet (2019), “information processing systems form a spectrum between two extremes: on one end, static systems that consist entirely of hard-coded priors… and on the opposite end, systems that incorporate very few priors and are almost entirely programmed via exposure to data,” while humans “combine ample amounts of both priors and experience” (p. 19). | Artificial intelligence (AI) refers to the ability of machines to perform tasks that typically require human-level intelligence, such as perception, reasoning, learning, and decision making. Human intelligence, on the other hand, is the ability of a person to think, understand, and learn from experiences and environment. AI operates based on pre-programmed algorithms and rules, whereas human intelligence is based on complex neural networks, emotions, biases, and past experiences. AI can process large amounts of data much faster than humans, but lacks creativity, empathy, and common sense that humans possess. AI can be programmed to make objective decisions, while human intelligence is subjective and influenced by emotions and personal experiences. |
| 4. How does “machine learning” differ from human learning? (~100 words) | |
| Machine learning is different from human learning because AI makes decisions based on patterns seen a limited data set without the ability to apply context or seek out new sources of information. Like humans, machines develop biases, but theirs are the result of data sets and variables from which they do not deviate (Heilweil, 2020). Janardhanan (2020) summarizes the major differences by stating that “an intelligent human being will be able to solve problems related to unforeseen domains, whereas a machine will not.” | Machine learning is a type of artificial intelligence that allows computer systems to learn from data without being explicitly programmed. Unlike human learning, machine learning algorithms don’t require a teacher or human intervention to change its behavior. Human learning is based on experience, perception, and feedback while machine learning relies on algorithms and large amounts of data to make predictions. Human learning can be influenced by emotions, biases, and past experiences, while machine learning is based solely on the data it was trained on. Both processes involve changing behavior over time, but machine learning is much faster and can process much larger amounts of data than human learning. |
| 5. How do your answers differ from what a machine could generate? Make sure you reference the specific differences between your answers and Chat GPT’s. (~200 words) | |
| Chat GPT did not cite any sources, whereas I was careful to only make claims I could back up with sources. For example, I know that the Imitation Game designed by Turing is commonly called the Turing Test, but since the paper I consulted did not mention that I made a conscious decision not to include the information. When I entered the questions into the Chat GPT textbox, I prefaced them with the number of words I expected in the answer (e.g., In 100 words, who was…). Chat GPT went over the word limit each time. I initially went over my word limit for each question but went back and deleted every word that I could without losing meaning or important information. In some instances, Chat GPT made claims that I was unable to confirm through internet searches, for example the statement about what Gebru believes to be the definition of intelligence. Unless she was quoted somewhere as saying those words, I think that the sentence is speculation and should have been worded as such. In comparison, my answer about her beliefs was less specific but can be traced back to a specific piece of writing. | |
References
BBC. (2016, January 26). Ai pioneer Marvin Minsky dies aged 88. BBC News. Retrieved January 31, 2023, from https://www.bbc.com/news/technology-35409119
Biography.com (Ed.). (2020, July 22). Alan Turing. Biography.com. Retrieved January 31, 2023, from https://www.biography.com/scientist/alan-turing
Chollet, F. (2019, November 5). On the measure of intelligence. Google, Inc. Retrieved January 28, 2023, from https://arxiv.org/pdf/1911.01547.pdf
Donovan, P. (n.d). Herbert A. Simon: Do we understand human behavior? The economics of altruism. Retrieved January 30, 2023, from https://www.ubs.com/microsites/nobel-perspectives/en/laureates/herbert-simon.html
Hao, K. (2020, December 4). We read the paper that forced Timnit Gebru out of google. here’s what it says. MIT Technology Review. Retrieved January 31, 2023, from https://www.technologyreview.com/2020/12/04/1013294/google-ai-ethics-research-paper-forced-out-timnit-gebru
Harris, A. (2018, November 1). Human languages vs. programming languages. Medium. Retrieved January 31, 2023, from https://medium.com/@anaharris/human-languages-vs-programming-languages-c89410f13252
Heilweil, R. (2020, February 18). Why algorithms can be racist and sexist. Vox. Retrieved January 31, 2023, from https://www.vox.com/recode/2020/2/18/21121286/algorithms-bias-discrimination-facial-recognition-transparency
Janardhanan, P. S. (2020, April 2). Human learning and machine learning – how they differ? Data Science Central. Retrieved January 31, 2023, from https://www.datasciencecentral.com/human-learning-and-machine-learning-how-they-differ/#:~:text=Let%20us%20examine%20the%20difference,the%20form%20of%20past%20data
Lee, J. A. N. (2013). Herbert A. Simon. Computer Pioneers – Herbert A. Simon. Retrieved January 30, 2023, from https://history.computer.org/pioneers/simon.html
McCarthy, J. (2007, November 12). What is artificial intelligence? Basic questions. Retrieved January 30, 2023, from http://www-formal.stanford.edu/jmc/whatisai/node1.html
Minsky, M. L. (1986). The society of mind. Simon and Schuster. Retrieved January 30, 2023, from https://archive.org/details/societyofmind00marv/page/17/mode/2up
OpenAI. (2023, January 25). CHATGPT: Optimizing language models for dialogue. OpenAI. Retrieved January 31, 2023, from https://openai.com/blog/chatgpt/
Turing, A. M. (1950). Computing, machinery and intelligence. Mind, 49(236), 433-460. Retrieved January 30, 2023, from https://www.cs.mcgill.ca/~dprecup/courses/AI/Materials/turing1950.pdf
Walsh, D. (2022, May 26). Timnit Gebru: Ethical AI requires institutional and structural change. Stanford University. Retrieved January 30, 2023, from https://hai.stanford.edu/news/timnit-gebru-ethical-ai-requires-institutional-and-structural-change