IP 2

Q 1.

1. Alan Turing was a gifted British mathematician who technically invented the first-ever computer. He laid out the theoretical frameworks for AI through publishing his 1950 paper, ‘’Computing Machinery and Intelligence’’. According to Turing, intelligence is identified as the ability to talk like humans or to pass as a human.

• John McCarthy was an American mathematician and computer scientist who pioneered AI. His contributions include creating programming languages that have a multitude of applications in several fields, and Timesharing systems which enable global information sharing. He identifies intelligence as the ability to map out the pathway to achieving goals.

• Herbert Simon was an American political scientist. He created the General Problem Solver program which simulates the human processes of decision-making and problem solving. According to Simon, intelligence is the ability to process and retrieve information to choose the course of action that leads to the best outcomes.

• Marvin Minsky was an American cognitive and computer scientist. He identified intelligence as having the human capacity for common-sense reasoning. His theoretical work, such as ‘’The Society of Mind’’ theory directed and informed research in AI. His book Perceptrons highlighted problems with neural networks which proved useful for the field.

• Timnit Gebru is an American computer scientist and AI ethics researcher. She helped diversify the field of AI in addition to revealing hard but important truths about the current state of the tech world. I think she  would define intelligence as the bravery to see injustice and speak up about it.

Q. 2

Human languages are heavily dependant on context; words can chage meanings based on context. In addition, there is space to be creative when using human languages by, for instance, inventing one’s own words that can be perfectly understood by the audience. Moreover, understanding human languages massively relies on taking into consideration non-verbal cues.

 On the other hand, in programming languages context is irrelevant; every unit or line of code means the same wherever it is used. Also, programming langauges are made up of very strict and clear rules; innovation has no place. Finally, non-verbal cues are nonexistent in programming languages.

Q 3.

AI is incredibly much faster at processing data than HI. However, HI has the advantage of being intuitive and able to act in unexpected situations. If anything derails off course, AI fails as it can only perform the tasks it was trained or programmed to perform. Moreover, AI cannot be biased as it acts based on facts only while HI can be opiniontaed and its judgement affected by personal agenda. Unlike AI, HI can lead to new inventions and discoveries. Furthermore, humans are faster at acquiring new skills. AI, in contrast, needs hundreds of years to learn the simplest skills.

Q 4.

Both machine and human learning depends on pattern recognition among large sets of data and model building. If data is not updated, machine learning becomes problematic as it runs the risk of algorithmic bias. Humans have the ability to draw inferences and apply or transfer knowledge to novel situations. Machines, however, can only interpret data inputted in them, not new concepts outside of their domain. Human learning is unique to each person while it is basically the same in all machines. There is a limited number of ways a machine can learn. In contrast, humans have many different learning styles.

Q 5.

My responses were generated based on prior experience, common sense, new knowledge, and my own analysis and synthesis of new knowledge learned. In the case of the machine, the responses would be based on past data already inputted by the designer which might be biased or inaccurate. Generating these responses have transformed my perspective on things, how I understand and will use them in the future. I have formulated my own way of seeing things and making sense of them. A machine cannot do that. It does not have a mind of its own. A machine’s response won’t be as specific as a human’s as it lacks the reasoning needed to choose exactly which information is of the maximum or most efficient relevance and how to best organize it.  However, a machine would definitely be much faster at finding the information required for these responses. But when it comes to selecting, organizing, and analyzing it as per the parameters of this assignment, a human would probably achieve a perfect result at this much faster than it would take the machine to learn how to do it successfully.