The world of automation has never been the same since the early large-scale language models (LLMs), Google's BERT and OpenAI's GPT-2, were released five years ago. These AI technologies and related technologies have penetrated deep into our online lives, and even among non-technical people, the word “algorithm” is commonly used to refer to the invisible hand of a more or less sophisticated person. It has become a phrase. A model that feeds our collective experience.
The release of ChatGPT towards the end of 2022 sparked widespread awareness and excitement about this type of model, impacting every aspect of our online lives. And marketers ask, “What does it mean to be human?” He asks two age-old philosophical questions. and “What is real?”
Here are some practical tips for business and marketing professionals to answer these questions.
Addressing concerns about the humanity and credibility of LLMs
we are all still human
No matter how connected we are, no matter how sophisticated companies' targeting, no matter how much data there is about us, we are still human.
For example, marketers may have a perception of who Gen Z “thinks” they are based on stereotypes or data they collect on their own. However, like every other generation, Gen Z is not a monolith and is made up of complex human beings with different beliefs and perspectives that are not necessarily reflected in their actions.
With so much uncertainty about the future, brands need to stay grounded in what's important to them on a real, human level, a level that LLMs can't calculate.
Dig deeper: Artificial Intelligence + Human Intelligence = Success
Avoid using LLM to generalize your audience or demographics.
This kind of overgeneralization, reducing groups to stereotypes, is not new. It's been around since the birth of marketing (or since the beginning of war, depending on how you look at it).
We have made great strides in moving away from LLM over the past few decades, but ironically, the easy access to LLM can set you back several steps if you are not careful.
For example, I asked ChatGPT, “What do Gen Z think about avoiding buying fast fashion?” And saw this output:
“Significant numbers of Gen Z actively advocate sustainable and ethical fashion practices. They often prefer to support brands that prioritize environmentally friendly and ethical production processes. Yes…please keep in mind that these are general trends and individual opinions within Gen Z may vary.”
Looking at this question from a different perspective based on consumer research, we asked the same question in our recent What Matters 2024 survey and found that 32% are currently avoiding fast fashion. This is the highest percentage of all generations surveyed. Yet, we find that two-thirds of Gen Z are not participating in sustainable fashion.
One of the great strengths of the LLM is its ability to produce 'plausible', coherent and well-written texts. They are not trained to provide correct, insightful answers, but provide answers based on mathematical averages of available information. When you use these models, you need to bring that acumen and acumen to them.
Dig deeper: Reduce the risk of generative AI by keeping humans in the loop
The importance of intimacy with consumers
Still, you should try to understand the people you want to connect with. You cannot assume that their individual wants and needs are included in the model's training set. You still need to interact with consumers, listen to them, and research them.
Once you understand your audience better, you can use that insight in conjunction with LLM and other tools to see your audience as a person, not just a group. This idea is known as consumer intimacy. It's more important than ever as marketers explore the capabilities of these new technologies.
Learn more: AI in Marketing: How to balance automation with the human touch
So how can you use these models most effectively?
Recent techniques such as search augmentation generation (RAG) and parameter efficient fine-tuning (PEFT) allow us to have the best of both worlds. Combined with the insights gained from real human understanding and the insights gained from stepping back and thinking in terms of the broader environment, these technologies can significantly enhance our efforts.
In addition to the power and flexibility of AI models, these models provide specific insights into the real-world behaviors and attitudes of the people you want to serve. Companies that can balance consumer intimacy with the incredible power of these new models will thrive in the coming years.
As we are already seeing, it can be a difficult journey, but sometimes the most important thing is to remember the obvious – we are all still human.
Learn more: MarTechBot updates: Personas, image generation, and new prompts
Get Martech! every day. free. It's in your inbox.
The opinions expressed in this article are those of the guest author and not necessarily MarTech. Staff authors are listed here.
