Week 5

Capstone Project Ideas

In the first post of this journal, I mentioned that I'm attending a college program for computer science. At the end of the program, we have to complete a large project, which is called the capstone project. For the capstone project we can work in groups or individually, and we can do a variety of things for the project. Some examples of projects include video games, video game modding software, data visualization software, and library software.

I have a few ideas for possible capstone projects I could do.

  • Develop a video game.
  • Develop a modding tool for a video game.
  • Develop an API or tool for a video game engine, such as Godot.
  • Develop a software that helps organize notes and ideas.

What are Stakeholders?

This week we read about stakeholders. A stakeholder is anybody that is impacted by a decision. An example of this would be a company dumping pollutants into a river or lake. The company is a stakeholder, as they are likely benefiting from dumping pollutants. Anyone who is in contact with the body of water is also a stakeholder, because they are being exposed to pollutants.

Stakeholders are an important consideration for ethical dilemmas. Analyzing the stakeholders of an ethical issue will tell you who is affected and why. It can help contextualize the issue, and help inform how to approach answering the issue.

Types of Claims

This week we also read about the different types of claims, and how to defend them. Claims are categorized into five different types: Claims of Cause, Claims of Definition, Claims of Fact, Claims of Policy, and Claims of Value.

Claims of Cause

A Claim of Cause asserts a cause and effect relationship between two events, that one event caused the other. An example would be "drunk driving increases the risk of being involved in an accident".

Claims of Definition

A Claim of Definition asserts the meaning, or definition, of a word or phrase. An example would be "A chihuahua is a breed of dog". Claims of definition are typically made with the assumption that the audience agrees with the claim.

Claims of Fact

A Claim of Fact asserts that something is true or false. An example would be "The earth orbits around the Sun".

Claims of Policy

A Claim of Policy asserts which policy or course of action should be taken. An example would be "The United States should implement Universal Basic Income".

Claims of Value 

A Claim of Value asserts that something is morally right or wrong. An example would be "Murder is wrong".

Uses of AI in Academic Writing

This week we read about the uses of Generative AI (GenAI) in academic writing. GenAI can help with academic writing by helping think, research, and write about topics. It is important to be careful about how GenAI is used for academic writing. This includes being mindful of the prompts you give GenAI, and understanding whether its output valid and relevant.

Failing to properly use GenAI in academic writing can be harmful. Using GenAI inappropriately can lead to plagiarism. Leaning to heavily on GenAI, as well as not being mindful of how it is used, can also be an obstacle to learning.

Prompt Engineering for GenAI

This week we read about prompt engineering. Prompt engineering is the process of carefully constructing a prompt to get a useful response from large language models and GenAI. There are four aspects to an effective prompt: Instruction, Context, Input Data, and Output Indicator.

  • Instructions are tasks that GenAI follows to produce its output.
  • Context is additional information that helps GenAI produce more accurate and relevant responses.
  • Input Data is the primary input or question for GenAI to base its output off of.
  • Output Indicator specifies how the GenAI's output should be formatted, such as a paragraph or sentence. 

 When using GenAI, it is important to engineer prompts based on the following criteria.

  • Introducing clear parameters and guidelines to avoid ambiguity in the output.
  • Removing biased and loaded language to avoid bias in the output.
  • Construct prompts so that they can encompass a larger scope to allow for a wider, more comprehensive output.
  • Include contextual clues in prompts to generate more accurate output.
  • Eliminate conflicting instructions from prompts to avoid output with unintended side effects.
  • Be mindful of the limitations of the GenAI to avoid inaccuracies.

Commenting on Other Journals

 This week we also left comments on some of our classmates' journals from last week.

Nasser's Blog: Concurrency

Suhaib's Blog: Week 4

 


 

 

Comments

Popular posts from this blog

Week 4

Week 2

Week 1