IP 2: Artificial Intelligence

Alan Turing

Alan was a British Mathematician who developed the first ‘computer’ or what came before the computer. He proved that not all mathematical problems had a definitive answer. He thought intelligence could be defined by how close AI can complete tasks similarly to a human.(Alan Turing – Education, Movie & Quotes, 2020) https://www.biography.com/scientists/alan-turing

Chat GPT:  Alan Turing (1912–1954), a British mathematician, contributed to artificial intelligence (AI) by proposing the Turing Test in 1950. He believed intelligence could be identified through observable behavior. If a machine’s conversation was indistinguishable from a human’s, it was considered intelligent. Turing’s pragmatic approach influenced early AI development, shaping natural language understanding. 

John McCarthy was an American computer scientist who coined the term ‘Artificial Intelligence.’ He believed intelligence was “the computational part of the ability to achieve goals in the world” (Sutton, 2020). Or in other words, using thinking and problem-solving skills to achieve goals.

 

Sutton, R. S. (2020). John McCarthy’s definition of intelligence. Journal of Artificial General Intelligence, 11(2), 66-67. 

Chat GPT: John McCarthy (1927–2011), an American computer scientist, coined “artificial intelligence” and played a pivotal role in AI’s formalization. He organized the 1956 Dartmouth Conference, marking AI’s inception. McCarthy identified intelligence through symbolic reasoning and problem-solving, shaping the development of AI systems with advanced cognitive abilities. 

Herb Simon was one of the founding fathers of artificial intelligence. His main area of interest was decisions making. He defined intelligence to be when computers solved problems like humans. 

Chat GPT: Herbert Simon (1916–2001), an American scholar, contributed to artificial intelligence (AI) by pioneering problem-solving and decision-making research. He developed the Logic Theorist, an early AI program. Simon identified intelligence through problem-solving capabilities, emphasizing symbolic reasoning. His work laid the foundations for AI systems with advanced cognitive abilities. 

Marvin Minsky was an American computer scientist and mathematician who contributed to the field of AI in robotics, mathematics, and cognitive science. He thought intelligence was based on common-sense reasoning and if systems could function with this human trait. 

Chat GPT: Marvin Minsky (1927–2016), an American cognitive scientist and computer scientist, co-founded the MIT Artificial Intelligence Project and made pioneering contributions to AI. He emphasized neural networks and perception in his work, believing that intelligence could be identified by emulating human cognitive functions, particularly in the realms of perception and learning. 

Gebru is an ethics researcher that focuses on biases and inequality in AI in areas such as facial recognition, energy use, and the impact on climate change. Although I could not find a definitive answer, I can assume she would value AI intelligence as one that can recognize and account for human diversity and aid in equality, or at least not support/exacerbate inequality. 

Chat GPT: Timnit Gebru is a computer scientist known for her work in artificial intelligence (AI) ethics and fairness. She co-founded the group Black in AI and has researched biases in AI algorithms. Gebru emphasizes the need for diverse perspectives in AI development to ensure fairness and avoid biases in intelligent systems. 

 Machine (programming) languages” vs human (natural) ones?  

How does “machine (artificial) intelligence” differ from the human version? 

How does “machine learning” differ from human learning?  

 Do YOUR answers to these questions differ from what a machine could generate?  

My answers to these questions differ from what a machine could generate in a few ways. Firstly, I used examples of experiences in my life to formulate answers and draw conclusions. I used my own assumptions and intuition. For example, when researching Timnit Gebru. I could not find her definition of intelligence and AI did not find me one either. I had to use my common sense and reasoning skills to come up with a potential definition. Additionally, I used references to articles to support my writing whereas AI does not have that capability (at least the version I am using). This makes the responses less specific and more general in nature. Lastly, the machine answers are better at writing more precisely and sound very professional, however the questions asked are more specific and factual requiring less opinions or problem-solving skills which are easier for AI to answer. 

1. Formulate a conception of usability (based on the Issa and Isaias (2015) chapter on HCI and Usability).  

HCI stands for Human-Computer Interaction, and it is a multidisciplinary field that focuses on the design and use of computer technology, emphasizing the relationship between humans and computers. Central to HCI is the concept of usability. This term includes many factors but essentially aims to create effective and user-friendly systems. This means that they should strive to be easily learned, navigated, individualized, memorable, and enjoyable for their users. To ensure usability, HCI should incorporate users in all stages of design.  

2. Then, think about what is missing from this conception, from a specifically educational perspective, and on that basis try and patch together a reasonably grounded and defensible conception of educational usability.  

The concept of usability mentioned above may not account for certain aspects of education. To start, usability as discussed by Issa & Isaias (2015), is often evaluated by speed, completion, and ease of use. However, in education evaluation of success surrounds learning concepts such as critical thinking and problem solving. These skills are not quick to develop and very difficult to assess. Additionally, learning is not meant to be easy. More traditional education such as memorization and regurgitation may fit with the concept of usability, however more contemporary education has shifted to collaboration and higher-level thinking. Educational usability would need to focus more on tools which promote learning, support diverse learners and educators, and have less concern for user satisfaction but rather users reaching their learning outcomes. In conclusion, when considering educational usability, a more comprehensive perspective is required compared to the conventional usability criteria. This entails considering the specific objectives and desired learning outcomes in the educational context. 

3. Revisit Woolgar’s (challenging but rewarding!) account of “usability gone wrong,” which demonstrates several ways a usability study ended up configuring ‘users,’ thereby undermining the usefulness of usability. Identify and discuss 2 of Woolgar’s examples. 

This paper uses examples from studying computers to explain how creating a new thing, like a new type of computer, involves shaping the people who use it. This process, called “configuring,” includes defining who the users are and putting limits on what they are likely to do in the future. By analyzing recordings of usability trials, where people test how easy it is to use the new machine, the paper suggests that deciding how well the machine and user relate to each other is influenced by the idea of setting boundaries between them. These boundaries can affect how users are perceived, how their actions are constrained, and how the machine is defined in relation to them. 

 Example 1  

In the story about the “wrong socket” incident, Ruth is trying to connect a machine to a printer. What stands out here is that the evaluation isn’t about Ruth’s abilities. Instead, it’s about checking if the instructions on how to use the machine effectively guide or shape the user, like Ruth, to complete the task correctly (Woolgar, 1990, pg 86). However, because the computer is the wrong type for the plug, we see how instructions and machines are shaping both Ruth and the observers when usability goes wrong. 

Example 2  

The manuals provide information for the user on what steps and actions they are to perform to use the machine. Therefore, “these captions configure the user by defining the correct courses of interpretation and action to be followed” (Woolgar 1990 pg 81). Because the machine was not complete the assessment involves considering whether the machine is behaving like a real machine and whether the user is acting like a real user in this evolving project. 

4. Finally, discuss the two excerpts quoted at the top of this IP, that have been drawn from your readings for this unit, and discuss differences you see in these 2 positions on the uses of usability. 

“…the usability evaluation stage is an effective method by which a software development team can establish the positive and negative aspects of its prototype releases, and make the required changes before the system is delivered to the target users”  (Issa & Isaias, 2015, p. 29). 

 

“…the design and production of a new entity…amounts to a process of configuring its user, where ‘configuring’ includes defining the identity of putative users, and setting constraints upon their likely future actions” (Woolgar, 1990). 

The difference between the 2 excerpts is how the user is viewed. In Issa and Isaia’s (2015) quote the user is utilized in the stages of development and assessment. They are the focus that the product will then change and evolve around for increased usability. Alternatively, Woolgar’s quote positions the user as something that will be configured to fit the machine. By placing constraints on the user, usability is determined by if the user can use the product within these set constraints. 

Sources:

Issa, T., & Isaias, P. (2015) Usability and human computer interaction (HCI). In Sustainable Design (pp. 19-35). Springer. 

Woolgar, S. (1990). Configuring the user: The case of usability trials. The Sociological Review, 38(1, Suppl.), S58-S99. 

TandR (2)