Things AI Can’t Do
Last week, I was encouraged by my media law professor to find areas, tasks, or results (names, writing a press release, creating a pitch presentation, market research, etc.) where AI doesn't seem to be doing a good job in the context of advertising and public relations. Here are some things I’ve learned after working with AI.
1). AI can’t review transcripts - transcripts are used in focus groups/interviews/ethnographies by AD/PR researchers. If you upload a transcript and ask for a thematic analysis or main points using only what you provided, it can't help itself and will make up quotes to match the trends it found
2). AI can’t compile tailored media lists - media lists are important to PR specialists when creating news coverage. However, AI’s responses are “ I'm not able to create lists based on personal opinion, but I can help you find some resources that might be helpful.” and “While I cannot provide a comprehensive list of 100 specific fashion editors and influencers in NYC due to data privacy concerns, I can offer you strategies to compile such a list yourself.” — AI cannot dive deeper into your brand’s authenticity and create a media lists based on your brand’s positioning among its categories
3). AI lacks recent data - AI systems often struggle with providing the most recent data, as they typically rely on information available up until their last update, therefore, they cannot deliver insights from the past year. Additionally, AI is unable to track or report on breaking news or current events as they unfold, further restricting its usefulness in situations where timely, accurate data or news coverage is essential.
4). AI cannot perform website audits - It does not have access to websites online and says "I can't access websites directly, but I can help you outline a detailed audit process for the "Keep Durham Beautiful" website. Here’s a structured approach to performing the audit" — it will then proceed to give general advice for any website. AI services like ChatGPT have limited resources to conduct actual research.
5). AI struggles with emotional nuance and cultural sensitivity - the system's ideas often come across as tone-deaf or lack empathy, missing the deeper emotional connection that resonates with audiences. This happened with Dove’s 2017 campaign where an AI-generated ad was criticized for being racially insensitive because it showed a black woman turning into a white woman after using a Dove product. While AI can analyze vast amounts of data to create campaigns, it often misses the subtle, human touch needed for messaging that’s culturally aware and emotionally attuned.
