Taxonomy of Delegated Tasks
Select specific tasks delegated to Artificial Intelligence.
Using AI to brainstorm new research topics, interdisciplinary connections, or novel angles on existing problems.
Refining vague goals into concrete, actionable research objectives.
Drafting and iterating on specific research questions (RQs) to ensure they are clear and answerable.
Assessing whether the proposed study is viable regarding resources, time, and data availability.
Conducting quick background checks or “pre-studies” to see if the idea has already been solved.
Using AI to conceptualize theoretical models, simulate persona responses, or design hypothetical scenarios.
Using AI tools (e.g., semantic search) to find relevant papers that keyword searches might miss.
Generating summaries or abstracts of long papers to quickly assess relevance.
Visualizing connections, citation networks, or thematic clusters in the literature.
Identifying trends or recurring themes across a large corpus of text.
Using AI to suggest areas where current research is lacking or contradictory.
Translating foreign-language literature to make it accessible for the review.
Designing the logic, control groups, and variables of the study.
Creating early drafts of survey instruments, interview guides, or experimental apparatus designs.
Suggesting appropriate statistical methods or qualitative frameworks for the data.
Generating boilerplate code, scripts, or functions from natural language descriptions.
Refactoring code for better performance or readability.
Identifying syntax errors or logical bugs in scripts.
Writing scripts to automate file management, backups, or batch processing.
Assisting in the logic and mathematical formulation of algorithms.
Generating docstrings, comments, and README files for research software.
Writing scrapers or using AI agents to gather public data.
Checking data for consistency, outliers, or errors.
Automating the formatting, parsing, and repair of messy datasets.
Organizing and categorizing large datasets.
Suggesting or performing statistical tests and interpreting raw outputs.
Generating code for plots, graphs, and data dashboards.
Summarizing data findings in textual or tabular format.
Using LLMs to annotate or classify text/image datasets (zero-shot/few-shot labeling).
Generating synthetic datasets to preserve privacy or augment small samples.
Detecting and removing Personally Identifiable Information (PII).
Translating textual data (e.g., survey responses) into the analysis language.
Creating conceptual images, illustrations, or diagrams from scratch.
Enhancing, cropping, or modifying existing images (e.g., removing background).
Creating flowcharts, process diagrams, or high-level infographics.
Generating initial text for sections based on bullet points or notes.
Correcting grammar, spelling, and punctuation.
Creating the abstract or plain-language summary.
Synthesizing the discussion into a final concluding statement.
Rewriting text to be more formal, concise, or accessible.
Translating the manuscript from the author's native language to the publication language.
Formatting citations and bibliography styles.
Drafting slide decks or conference poster text.
Brainstorming catchy and accurate titles for the work.
Scanning text or study designs for potential cultural or gender bias.
Identifying potential dual-use concerns or societal risks.
Checking against specific ethical guidelines or checklists.
Ensuring no private data is inadvertently leaked in the text.
Using AI as a reviewer to identify unclear sections or potential critiques.
Checking the manuscript against current research trends and ensuring internal consistency.
Identifying weaknesses or limitations in the study that the authors missed.
Suggesting suitable journals or conferences and helping with the submission strategy.
Assisting in drafting cover letters, responses to reviewers, or other publication-related materials.