IP 2: Artificial Intelligence

Question	My Answer	Chat GPT Answer
Alan Turing	Turing, an English mathematician, laid the foundation for AI by constructing a model for human computation and working to prove whether or not machines could imitate it, creating the Turing Test to identify success and failure (Muggleton, 2014; Turing, 1950).	Alan Turing was a British mathematician and computer scientist. He played a pivotal role in AI by conceiving the Turing Test, a method to identify machine intelligence based on a computer's ability to mimic human conversation convincingly. His work laid the foundation for modern AI and computing.
John McCarthy	McCarthy was an American computer scientist who coined the term “artificial intelligence” in a proposal he co-wrote and presented at the foundational Dartmouth Conference (“John McCarthy,” 2023). For him, intelligence was present in any organism or machine that had beliefs, indicated by an ability to problem-solve (“John McCarthy,” 2023).	John McCarthy was an American computer scientist and AI pioneer. He coined the term "artificial intelligence" (AI) and developed the Lisp programming language, a key tool for AI research. McCarthy believed that intelligence could be identified through problem-solving and reasoning abilities in machines, shaping AI's early goals.
Herb Simon	Simon was an American political scientist whose work on decision making and cognition has heavily influenced artificial intelligence (“Herbert A. Simon,” 2023). For Simon, an intelligent machine needs to hypothesize and test during problem-solving, which could be assessed using protocol analysis. (Dobrin, 1986).	Herbert Simon was an American psychologist and computer scientist. He co-developed the logic theorist program, a pioneer in AI, and proposed the "bounded rationality" concept, suggesting that intelligent behavior could be identified by observing how well a system adapts to its environment within limited cognitive resources, influencing AI research.
Marvin Minksy	Minsky founded a number of models of artificial intelligence, some controversial, which continue to influence AI research today (“Marvin Minksy,” 2023). Minsky believed an intelligent machine was one that could complete any task a human could do (Fjelland, 2020).	Marvin Minsky was an American cognitive scientist and AI pioneer. He co-founded the MIT AI Lab and made significant contributions to robotics and computer vision. Minsky believed intelligence could be identified by breaking down complex tasks into simpler processes, inspiring research in AI problem-solving and knowledge representation.
Timnit Gebru	Gebru is an Eritrean Ethiopian computer scientist and AI researcher renowned for her expertise in ethical and inclusive artificial intelligence ("Timnit Gebru," 2023). According to Gebru, AI should be expected to create without doing harm to populations or the environment ("Timnit Gebru," 2023; Small, 2023).	Timnit Gebru is an Ethiopian-American computer scientist renowned for her AI ethics research. She co-authored influential work on bias and fairness in AI algorithms. Gebru emphasizes the importance of addressing societal and ethical aspects in AI development, believing that intelligence identification should involve responsible, unbiased, and equitable AI systems.
How do “machine (programming) languages” differ from human (natural) ones?	Machine languages and human ones differ in their degree of performativity, with machine languages being entirely performative and natural languages requiring certain contexts and conditions to be so (Jones, 2020). Additionally, Human language includes layers of clues, which is particularly clear when speech is considered, and is both logical and emotional (Harris, 2018). Programming languages are designed in their entirety in the outset, whereas human languages develop and evolve over time (Harris, 2018).	Machine or programming languages differ from human languages in several ways. First, they are designed for specific, unambiguous instructions to computers, lacking the richness, ambiguity, and context of human languages. Second, they have strict syntax and grammar rules, with no room for interpretation. Human languages are dynamic and context-dependent, with nuances and cultural variations. Additionally, machine languages are typically binary or low-level, while human languages are high-level and expressive, allowing for creativity, emotion, and abstraction. Finally, humans use language for communication and expression, while machines use it to execute tasks, making their purposes fundamentally distinct.
How does “machine (artificial) intelligence” differ from the human version?	Definitions of human intelligence have varied over the centuries and even decades of the most recent century (Chollet, 2019; Crawford, 2021). As such, a definition of artificial intelligence has varied as well. However, increasingly it’s clearn that human intelligence has a high degree of generality and openendness not seen in current intierations of machine learning, which still views human intelligence as a blank slate upon which networks of process and data retrieval are built to ultimately efficiently solve problems (Chollet, 2019). Machine learning uses correlation to acquire knowledge (Jones, 2020) and requires explicit, formalized processes and data (Crawford, 2021). Crawford states human intelligence is also situated within “social, cultural, historical and political” contexts not present for AI (2021, p. 5).	Machine or artificial intelligence (AI) differs from human intelligence in several key ways. Firstly, AI lacks the depth and breadth of human understanding, as it relies on algorithms and data rather than true comprehension. AI lacks common sense reasoning and the ability to understand context and emotions as humans do. Human intelligence also encompasses creativity, empathy, and consciousness, which AI currently lacks. Moreover, human intelligence is adaptable and can learn a wide range of tasks, while AI is often specialized and requires extensive training. Additionally, human intelligence is inherently tied to subjective experience, which AI cannot replicate.
How does “machine learning” differ from human learning?	Machine learning uses correlation to acquire knowledge (Jones, 2020) and requires explicit, formalized processes and data (Crawford, 2021). Humans can use both correlation and causation and have a open, generalized way of constructing knowledge (Jones, 2020). Humans are born wired with prior knowledge that experience builds upon and influences within a rich context of culture and environment, whereas a machine’s learning system is built upon a blank slate even if it does evolve its knowledge bank over time (Chollet, 2019; Crawford, 2021).	Machine learning and human learning differ fundamentally in their processes and capabilities. Machine learning relies on algorithms and data to improve performance on specific tasks through pattern recognition and statistical optimization. It lacks human-like understanding, common sense reasoning, and the ability to generalize knowledge across diverse domains. Human learning is a complex, cognitive process involving sensory perception, memory, emotion, and critical thinking, leading to a deep understanding of the world, creativity, and adaptability. Human learning also includes social and cultural aspects that shape knowledge acquisition, while machine learning is solely data-driven. The two processes serve different purposes and operate on distinct principles.
Comparing my responses to ChatGPT's.	Each of my answers differed in some way from Chat GPT’s answers. When writing 50 words about each of the individuals profiled above, I found it difficult to learn enough about their impacts and views on intelligence to summarize it into 50 words. What seemed worth retaining or mentioning explicitly didn’t always match ChatGPT’s choices. For instance, ChatGPT included the Lisp programming language in McCarthy’s profile but I hadn’t realized that Lisp specifically related to AI and thought it was a tangential achievement. I had been interested to learn how Minksy’s models of AI were controversial to some but this issue was apparently not enough to warrant a mention in ChatGPT’s profile. ChatGPT described Minksy in a much more celebratory and authoritative light. Generally, I struggled to digest all the information I was reading into the word count and often my text is disjointed. ChatGPT is much more eloquent in the space allotted, as is particularly apparent in the Gebru profile and the final three questions. Of course, a considerable difference is that ChatGPT cites no sources. Though outside the scope of this question, it is worth noting that ChatGPT completed this task incredibly fast compared to me researching, reading, writing and editing.

References

Chollet, F. (2019, November 5). On the measure of intelligence. Links to an external site.

Crawford, K. (2021). Atlas of AI. Yale University Press. (Introduction: pp. 1-21)

Dobrin, D. (1986). Protocols once more. College English, 48(7), 713-725. https://www.jstor.org/stable/377377

Fjelland, R. (2020). Why general artificial intelligence will not be realized. Humanities and Social Sciences Communications, 7, 1-9. https://www.nature.com/articles/s41599-020-0494-4

Harris, A. (2018, October 31). Human languages vs. programming languages. In Medium. https://medium.com/@anaharris/human-languages-vs-programming-languages-c89410f13252

Herbert A. Simon. (2023, September 20). In Wikipedia. https://en.wikipedia.org/wiki/Herbert_A._Simon#Pedagogy

John McCarthy. (2023, August 8). In Wikipedia. https://en.wikipedia.org/wiki/John_McCarthy_(computer_scientist)#Philosophy_of_artificial_intelligence

Jones, R. H. (2020). The rise of the Pragmatic Web: Implications for rethinking meaning and interaction. In C. Tagg & M. Evans (Eds.), Message and medium: English language practices across old and new media (pp. 17-37). De Gruyter Mouton.

Marvin Minksy. (2023, September 5). In Wikipedia. https://en.wikipedia.org/wiki/Marvin_Minsky

Small, T. (2023, May 15). The downside of AI: Former Google scientist Timnit Gebru warns of the technology’s built-in bias. The Globe and Mail. https://www.theglobeandmail.com/business/article-the-downside-of-ai-former-google-scientist-timnit-gebru-warns-of-the/

Timnit Gebru. (2023, August 12). In Wikipedia. https://en.wikipedia.org/wiki/Timnit_Gebru