Developers using AI showed the same core problem-solving behaviors as those without but differed in how they became stuck and recovered, with AI helping or hindering in specific cases.
It’s Weird That it Knows What I Want
9 Pith papers cite this work. Polarity classification is still indexing.
citation-role summary
citation-polarity summary
verdicts
UNVERDICTED 9roles
background 4representative citing papers
Large-scale analysis of wild LLM chat logs finds that user interaction patterns stabilize quickly after initial use and correlate with long-term outcomes like retention, creating an agency paradox of limited exploration in unconstrained systems.
EvoGraph turns linear AI-assisted programming into a manipulable graph of branching histories, reducing cognitive load and enabling better iteration according to a user study with 20 developers.
Novices performed better and reported lower workload with GitHub Copilot than with human partners, but human partners produced more positive emotions and a smaller drop in retest performance after one week.
Developers using AI assistants exhibit more stable emotions and greater focus on code creation, evaluation, and verification, captured in a new four-dimensional S-IASE model from retrospective labeling of screen recordings, surveys, and interviews.
EditFlow reconstructs temporal developer editing flows from code changes to benchmark and optimize AI code edit recommenders so they align with natural incremental reasoning rather than static snapshots.
Among novice programmers using AI code generators, trust did not predict compliance with suggestions, while performance correlated with both compliance and increased subsequent trust.
An experience report describes the design and delivery of an integrated CS1-plus-Discrete-Structures studio course called CS 1.5 that treats AI as a partner and emphasizes theoretical foundations through code comprehension projects.
A survey of user studies on LLM use in programming that identifies interaction behaviors, mixed benefits and weaknesses, and factors influencing human and task performance.
citing papers explorer
-
ChatGPT: Friend or Foe When Comprehending and Changing Unfamiliar Code
Developers using AI showed the same core problem-solving behaviors as those without but differed in how they became stuck and recovered, with AI helping or hindering in specific cases.
-
Priming, Path-dependence, and Plasticity: Understanding the molding of user-LLM interaction and its implications from (many) chat logs in the wild
Large-scale analysis of wild LLM chat logs finds that user interaction patterns stabilize quickly after initial use and correlate with long-term outcomes like retention, creating an agency paradox of limited exploration in unconstrained systems.
-
Choose Your Own Adventure: Non-Linear AI-Assisted Programming with EvoGraph
EvoGraph turns linear AI-assisted programming into a manipulable graph of branching histories, reducing cognitive load and enabling better iteration according to a user study with 20 developers.
-
Fast and Forgettable: A Controlled Study of Novices' Performance, Learning, Workload, and Emotion in AI-Assisted and Human Pair Programming Paradigms
Novices performed better and reported lower workload with GitHub Copilot than with human partners, but human partners produced more positive emotions and a smaller drop in retest performance after one week.
-
How Do Developers Interact with AI? An Exploratory Study on Modeling Developer Programming Behavior
Developers using AI assistants exhibit more stable emotions and greater focus on code creation, evaluation, and verification, captured in a new four-dimensional S-IASE model from retrospective labeling of screen recordings, surveys, and interviews.
-
EditFlow: Benchmarking and Optimizing Code Edit Recommendation Systems via Reconstruction of Developer Flows
EditFlow reconstructs temporal developer editing flows from code changes to benchmark and optimize AI code edit recommenders so they align with natural incremental reasoning rather than static snapshots.
-
Relationships Between Trust, Compliance, and Performance for Novice Programmers Using AI Code Generation
Among novice programmers using AI code generators, trust did not predict compliance with suggestions, while performance correlated with both compliance and increased subsequent trust.
-
"CS 1.5": An Experience Report on Integrating CS1 and Discrete Structures for the AI Era
An experience report describes the design and delivery of an integrated CS1-plus-Discrete-Structures studio course called CS 1.5 that treats AI as a partner and emphasizes theoretical foundations through code comprehension projects.
-
Understanding the Human-LLM Dynamic: A Literature Survey of LLM Use in Programming Tasks
A survey of user studies on LLM use in programming that identifies interaction behaviors, mixed benefits and weaknesses, and factors influencing human and task performance.