Looking back at AI and qualitative research in 2025

Happy Holidays

As we close out 2025 and head into 2026, it’s clear: AI adoption kept growing — and so did the resistance to it.

On the adoption side, more teams moved from “trying AI” to building it into everyday work. On the resistance side, we saw a more explicit pushback: for example, an open letter by hundreds of experienced qualitative researchers rejecting the use of GenAI for “Big Q” approaches such as reflexive thematic analysis.

In parallel, “slow” movements gained visibility in art, culture, and research settings — including events and initiatives explicitly framed as Slow AI.

That mix isn’t surprising. Fast innovation brings uneven outcomes: real productivity gains for some, real harm (or fear of harm) for others, and a lot of confusion in the middle. Resistance is part of the process of society negotiating what counts as acceptable use, credible evidence, and responsible practice.

A Word about GPT 5 As an indicator how fast-paced the development is, remember that just a few weeks ago, I told you that I was no longer happy with the OpenAI chatbot after the release of version 5 and 5.1.

The update to GPT-5.2 brought performance back to a level I’m genuinely happy with. I am still using Gemini 3 pro. It depends on the task at hand. At times I also check what suggestion I get from both models - and then choose the best outcome. This is not for data analysis (for this I would not use chatbots), but for general every day tasks like finding a better wording for a critical email, or writing a social media posts on the ideas that I outline.

Key Developments in 2025 • Rise of Autonomous AI Agents: AI agents have evolved from simple chatbots to sophisticated, autonomous systems capable of planning and executing complex, multi-step workflows without constant human oversight. • Advancements in AI Reasoning and Efficiency: Progress is marked by better "reasoning" capabilities in AI models, which allow them to break down problems and execute multi-step tasks more effectively. • Multimodal AI Takes Center Stage: AI models can now seamlessly process and generate content across various formats, including text, images, audio, video, and even 3D content. • Major Impact on Healthcare and Science: AI is accelerating R&D significantly. this includes Drug Discovery and Diagnostic: AI models are simulating molecular interactions to speed up the discovery of new drugs and treatments for diseases like cancer and genetic disorders. AI tools are being used for early and highly accurate detection of conditions such as cardiovascular disease, dementia, and even predicting biological aging from standard medical scans like EKGs and X-rays. • Increased Focus on Ethics and Regulation: With the rapid proliferation of AI, there is a renewed emphasis on building safe, secure, and transparent systems. Efforts include developing robust testing protocols, implementing embedded watermarking for AI-generated content to combat misinformation.