Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
38 tokens/sec
GPT-4o
59 tokens/sec
Gemini 2.5 Pro Pro
41 tokens/sec
o3 Pro
7 tokens/sec
GPT-4.1 Pro
50 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Prompt Problems: A New Programming Exercise for the Generative AI Era (2311.05943v1)

Published 10 Nov 2023 in cs.HC

Abstract: LLMs are revolutionizing the field of computing education with their powerful code-generating capabilities. Traditional pedagogical practices have focused on code writing tasks, but there is now a shift in importance towards code reading, comprehension and evaluation of LLM-generated code. Alongside this shift, an important new skill is emerging -- the ability to solve programming tasks by constructing good prompts for code-generating models. In this work we introduce a new type of programming exercise to hone this nascent skill: 'Prompt Problems'. Prompt Problems are designed to help students learn how to write effective prompts for AI code generators. A student solves a Prompt Problem by crafting a natural language prompt which, when provided as input to an LLM, outputs code that successfully solves a specified programming task. We also present a new web-based tool called Promptly which hosts a repository of Prompt Problems and supports the automated evaluation of prompt-generated code. We deploy Promptly for the first time in one CS1 and one CS2 course and describe our experiences, which include student perceptions of this new type of activity and their interactions with the tool. We find that students are enthusiastic about Prompt Problems, and appreciate how the problems engage their computational thinking skills and expose them to new programming constructs. We discuss ideas for the future development of new variations of Prompt Problems, and the need to carefully study their integration into classroom practice.

An Examination of "Prompt Problems: A New Programming Exercise for the Generative AI Era"

The paper "Prompt Problems: A New Programming Exercise for the Generative AI Era" details an innovative approach to programming education, aiming to refine the increasingly important skill of prompt engineering when interacting with Generative AI models to produce code solutions. Prepared by leading academics in the field of computing education, the work demonstrates a shift from traditional code writing pedagogies to emphasizing code comprehension and prompt crafting as essential components of modern programming instruction.

The authors introduce "Prompt Problems", a type of programming exercise focused on generating code through constructing effective natural language prompts for LLMs. As educational paradigms adapt to the capabilities offered by LLMs, the ability to generate meaningful and precise prompts becomes crucial for students. To facilitate this learning process, the researchers presented a platform named "Promptly", which hosts these exercises and evaluates prompt-generated code through test cases. Findings from a paper with two courses demonstrate that students engage positively with these exercises, appreciating the challenges they present in honing computational thinking and programming skills.

Quantitative analysis within the paper assesses the effectiveness and efficiency of prompts crafted by students for solving given problems. Results reveal a broad variation in prompt quality and length, yet many students successfully solve exercises within a limited number of attempts. The average prompt length and number of submissions required for success provide valuable insights into problem comprehension and engagement, revealing that students tend to solve problems after a few submissions. Notably, intricate problem-solving tasks prompt students to explore novel coding constructs, enhancing their programming acumen.

Feedback from participants underscores an appreciation for the exposure to new programming constructs and the mental exercises involved in understanding and generating prompts. These exercises, in the view of many students, contribute to developing an analytical skill set aligned with the demands of modern programming environments where AI-generated solutions play a pivotal role. Some students voiced concerns over potential over-reliance on AI-driven coding and expressed anxiety about automation's impact on programming careers, articulating a need for balanced integration in the curriculum.

This work provides significant implications for computer science education. The introduction of prompt crafting as a fundamental concept could transform instructional practices, demanding educators reassess curriculum structures to incorporate skills relevant to AI interactions. While traditional coding skills remain essential, proficient prompt engineering offers a complementary skill set that reflects evolving industry expectations. Future developments may focus on exploring alternative implementations of "Prompt Problems", such as structured prompts versus exploratory dialogue prompts, analyzing optimal integration strategies, and expanding exercises to other programming paradigms.

Overall, the paper enriches the discussion surrounding AI-enabled education tools, presenting a credible validation for embedding prompt engineering as a vital component of coding curricula. As generative AI becomes increasingly integrated into technological solutions, educational methods must evolve in tandem, cultivating a workforce adept at harnessing AI for innovative problem solving.

This essay adopts an academic tone appropriate for an audience of experienced researchers, providing detailed insights into the implications and methodologies explored in the paper, while speculating on future directions for AI developments in computing education.

User Edit Pencil Streamline Icon: https://streamlinehq.com
Authors (7)
  1. Paul Denny (67 papers)
  2. Juho Leinonen (41 papers)
  3. James Prather (21 papers)
  4. Andrew Luxton-Reilly (16 papers)
  5. Thezyrie Amarouche (4 papers)
  6. Brett A. Becker (14 papers)
  7. Brent N. Reeves (9 papers)
Citations (49)
Youtube Logo Streamline Icon: https://streamlinehq.com