Techniques

Applied AI

Unprecedented Innovation Creating Opportunity, Uncertainty, and Risk

AI is spawning a dizzying array of platforms and tools with significant use potential in the insights industry, although most of these applications are currently based on Large Language Models (LLMs) rather than computational platforms. LLMs are, by design, not good at math. But they are capable of calling up and giving instructions to computing tools, and eventually, they will get good at interpreting and describing the results.

There is great variation in the quality and sophistication of AI products for the insights industry and ongoing uncertainty about ideal use cases. Adding to our challenges is the critical need to ensure that we don’t use AI to automate new sources of error into our work. All of us know the risk of shiny objects. To get it right, we need to explore this new AI world imaginatively but also drive defensively.

We’re taking our historic stewardship role seriously, balancing active experimentation with careful assessment and strategic customization to get the best out of AI.

Careful Curation for Optimal Application

NAXION’s multidisciplinary innovation team, spearheaded by behavioral and data scientists, is systematically investigating potential AI applications and testing a host of new AI products. We’re collecting feedback and contributing to the conversation.  It’s an important one to be having.

A critical part of the curation process is formalizing the QC steps needed to adopt and supervise AI tools without trading accuracy for speed. Humans are still very much in the loop.

High on the priority list for evaluation at NAXION are LLM-based tools designed to make qualitative analysis more efficient and widgets that can be incorporated into surveys for richer probing of open-ends. We’re also keeping a close and cautious eye on qual-at-scale, where the risk-reward calculation is far more complex.

On the quantitative side, we’re test-driving platforms that can be programmed to guide statistical analysis by calling up relevant techniques. And we’re actively building programs to enhance the use of off-the-shelf products to improve and customize current offerings.

Many of the new tools are quite powerful, but the near-term impact appears to be primarily on operational efficiency. They are not yet performing tasks humans can’t already do, but by taking over certain routinized tasks, they are giving our teams more time to do the things computers can’t.

Keep an Eye on This Space

As we continue to gain perspectives on AI, we will be publishing Fresh Thinking articles to share our experiences and recommendations. Please keep an eye on this space for regular updates and links to research-on-research results and practical guidance to help navigate this dynamic landscape. And if you want to chat with a human at NAXION about this or any other topic, we’re standing by.

Fresh Thinking