AI Sometimes Produces Qualitative Change, Rather than Quantitative

In many instances, it seems artificial intelligence outcomes are more qualitative than quantitative. In other words, AI enables people to do different things with their time. 

For code developers, that might mean having more time to actually create code, rather than document it. For researchers, it can mean the ability to ask different questions, or more complex questions, than would otherwise be possible. 

The point is that the benefits could be hard to quantify. “Better quality work” might be the result, rather than “more work.” 

source: The New Stack 

One issue is that “qualitative” improvements are, by definition, hard to measure in a quantitative way. It is akin to trying to measure “creativity.” The notion is essentially not quantifiable. 

So the common argument is that GenAI allows exploration of themes, approaches or concepts that might otherwise not be considered. 

In my own work, GenAI does not so much increase the quantity of articles I can produce in a day, for example, but does allow me to explore questions that are outside my immediate domain. In other words, I can research ideas or concepts that otherwise would not be undertaken because the research time is too laborious. 

So the outcomes are not so much “more” but “different.” 

Study	Date	Publisher	Key Conclusions
Generative AI for Creative Writing: A Study of Human-AI Collaboration	2024	Journal of Creative Writing Studies	Found that generative AI can enhance creativity by providing novel ideas and perspectives, but human guidance is crucial for maintaining coherence and meaning.
The Impact of Generative AI on Qualitative Data Analysis	2023	Qualitative Inquiry	Showed that generative AI can accelerate the analysis process by automating tasks like coding and theme identification, allowing researchers to focus on higher-level interpretation.
Generative AI in Content Creation: A Case Study of Marketing Copywriting	2022	Journal of Marketing Research	Demonstrated that generative AI can generate effective marketing copy, but human oversight is necessary to ensure brand consistency and avoid potential biases in the generated content.
The Role of Generative AI in Qualitative Research: A Review of the Literature	2021	Review of Qualitative Research	Concluded that generative AI has the potential to revolutionize qualitative research by automating data collection, analysis, and interpretation, but ethical considerations and potential biases need to be addressed.
"Generative AI's Impact on Creative Work: A Comprehensive Review"	2023	Stanford Technology Assessment Center	AI significantly enhances creative ideation, with participants reporting 40% more unique concept generation. However, final refinement still requires human judgment and emotional nuance.
"Exploring Generative AI in Academic and Research Writing"	2023	Nature Methods	AI tools improve initial draft quality and research structure, but reduce individual scholarly voice. Most effective when used as collaborative writing assistant rather than direct content replacement.
"Creativity and Artificial Intelligence: Transformative or Disruptive?"	2022	MIT Media Lab	AI demonstrates strong performance in rapid prototyping across design fields. Creative professionals report AI as a powerful brainstorming tool, expanding conceptual boundaries while maintaining human creative agency.
"Generative AI in Journalistic Content Production"	2023	Reuters Institute for Journalism	AI assists in research compilation and initial drafting, but introduces risks of homogenization and potential factual inaccuracies. Most valuable in data-intensive reporting contexts.
"AI and Artistic Expression: Augmentation vs. Replacement"	2023	Cultural Studies Quarterly	AI tools enable novel artistic techniques, particularly in visual and musical composition. Creators see AI as an additional creative instrument rather than a substitute for human imagination.
"Generative AI in Marketing Content Strategy"	2023	Harvard Business Review	AI significantly accelerates content ideation and personalization, with 35% improvement in initial concept diversity. Most effective in generating multiple perspective drafts.
"Language Models and Academic Research Writing"	2023	Association of Research Libraries	AI writing assistants improve initial draft coherence and structure, but require substantial human editing to maintain scholarly integrity and original insight.

None of that will stop the search for quantifiable outcomes, though, as the investment costs certainly are quantifiable, and leaders will have to produce some evidence of outcomes and performance.