Meghan Surdenas wrote an article for Quirks!! #canAIdothat
If you’ve been to any market research conference or symposium, you’ve heard a lot of information on AI. It has transformed how some people work, streamlined many processes and aided some administrative tasks; but does anyone remember blockchain? The environment surrounding this technology is eerily similar: exponential growth, limited oversight and governance, a fundamental lack of accountability and a whole lot of industry buzz.
Is AI the new blockchain?
“AI won’t take your job – if you know how to use it,” economist Richard Baldwin said at the World Economic Forum’s Growth Summit. Fundamentally, AI adoption is simple. It’s technology that can be used to your advantage, as another tool in your toolbox. However, it is crucial to recognize the limitations of AI in research and how it falls woefully short in delivering the essential human touch. The value of researchers and research support personnel cannot easily be replaced. People still matter.
Qualitative recruitment: The new online dating
At its core, a qualitative project seeks to deeply understand the human condition and is far more complex than simply identifying suitable respondents based on objective criteria. Remember dating in high school and beyond? A person may look great on paper (or online these days), but an expert recruiting firm has people-centered processes that ensure respondent selection meets project goals with respondents who are articulate and engaging. It involves understanding individuals’ unique experiences, motivations and interpersonal skills where people can make informed decisions based on comprehensive insights beyond surface-level data. AI algorithms struggle to decipher subtle nuances, nonverbal cues and emotional intelligence that are vital to research quality standards.
Successful qualitative recruitment requires boots on the ground to ensure respondents will offer the insight for a successful outcome. Say goodbye to recruiting from the database. Recruiters need to get creative when needed, networking, researching, utilizing partners and old-fashioned handshakes. Difficult recruitment projects can be a puzzle and AI isn’t able to handle the nuances of that yet.
Transcription: The Human Touch
AI transcription isn’t new technology and is primarily designed for consumer interviews with common English. While AI does a decent job recognizing words and generating verbatim transcripts, it often struggles with different accents, dialects, contextual nuances and multiple speakers. AI systems may misinterpret homonyms, idiomatic expressions and technical or medical jargon, leading to inaccuracies and misrepresentations.
There is no substitute for human transcripts. Need the sleek sophistication of an Excel transcript? Good luck with AI. Client-specific formatting? AI can’t do it. That notwithstanding, there is also a growing need for a human review to create a client-ready transcript: looking up industry specific terms, following detailed delivery and formatting instructions. Human transcriptionists, with their contextual knowledge, linguistic expertise and cultural sensitivity, can better decipher these nuances and provide accurate and meaningful transcripts.
Data analysis: Trust But Verify
AI algorithms can be a standout in synthesizing structured data and identifying patterns but fall short in interpretation with unstructured information. Human analysts possess the ability to delve deeper into the data and provide valuable insights that go beyond statistical correlations. They can identify hidden patterns, discern underlying consumer behaviors and offer a more holistic understanding of market dynamics, which aids in strategic decision-making.
Project teams rely on real experience to provide valuable context to the data, considering factors such as macroeconomic conditions, competitive landscapes and consumer sentiment, which is crucial in generating accurate and actionable insights.
Here’s a secret: This article was initially generated by AI. It was excruciatingly elementary, devoid of any finesse. Given that, AI did generate the kernel of an idea, which led to collaboration with colleagues. Full disclosure? My longest-standing colleague in research generated the real message. People and relationships still matter. #canAIdothat
See the article in Quirks: https://www.quirks.com/articles/back-to-basics-what-ai-doesn-t-do-for-qualitative-support-services