Connect with us:
Connect with us:
Once thought to be the stuff of science fiction, AI (Artificial Intelligence) is increasingly having a major impact on business, especially marketing. While some organizations are integrating new Generative AI tools into their marketing initiatives as soon as they come online, many marketers are still trying to understand how to use these new tools.
At Grey Matter Direct, located outside of Philadelphia in Mount Laurel, New Jersey, we are embracing new Generative AI tools while also studying the implications that these powerful new tools will have on our client’s long-term marketing objectives.
Traditional Artificial Intelligence (also referred to as Narrow or Weak AI) is a subset of artificial intelligence that is used to perform tasks based on predetermined algorithms. For the most part, traditional AI solutions are designed to carry out a focused set of tasks, being more restricted in the scope of potential applications than other types of AI.
Some examples of traditional AI include applications like chatbots, autonomous vehicles, or chess-playing algorithms.
Generative AI is an artificial intelligence that is capable of generating outputs such as text, images, and other data. This type of AI works primarily by taking large quantities of existing data, analyzing it, and based on those findings, producing new content. What this means is that generative AI relies on machine learning to recognize, predict, and create content using the data sets it can access.
A large language model (LLM) is a type of artificial intelligence (AI) program that can recognize and generate text, among other tasks. LLMs are trained on huge sets of data — hence the name “large.” LLMs are built on machine learning: specifically, a type of neural network called a transformer model.
Many LLMs are trained on data that has been gathered from the Internet — thousands or millions of gigabytes’ worth of text. But the quality of the samples impacts how well LLMs will learn natural language, so an LLM’s programmers may use a more curated data set.
Although AI data scientists are not exactly sure how the LLMs work, they seem to gain a degree of “intelligence” through the training set data and are able to construct highly competent text. AIs are even being used to create entire books.
The concern is that there are now billions of web pages being produced by Large Language Models. The volume of this new content is starting to overwhelm Google’s ability to absorb all of the pages and properly index them for search. As a result, Google and the other search engines are looking for ways to determine if text is AI generated or not and the tools to sniff out the digital signals of this content are getting more sophisticated all the time.
Since Google’s core function is to present relevant and authentic content to users, the search giant will be trying to punish pure AI content and reward human-generated content. Because of this, we warn clients not to rely too heavily on ChatGPT, Grok, Gemini, Claude or other AI tools to produce written content since your site’s SEO could be downgraded in the future. Instead, we recommend using these tools to organize your thoughts into narratives and to gather raw information. Then, humans should heavily edit and rewrite this content to make it their own.
Occasionally, AI systems will present you with completely incorrect information. AI scientists call these “hallucinations”. We prefer to call them lies. A great example of this is when we recently asked ChatGPT for the biography of one of our clients, the former Philadelphia Flyers enforcer, Todd Fedoruk. ChatGPT confidently produced his bio, but it was completely wrong. When we told ChatGPT that it was wrong, the system said “Sorry about that. Here is his correct bio.”
It is unclear why the system would produce bad information since it is trained on internet sources and Todd’s bio is clearly laid out on his Wikipedia page.
These situations should give every marketer pause when using LLMs to produce content. You can use these systems but make sure you double check the content they produce.
Another compelling use for Generative AI is image generation.
Tools like DALL-E and Midjourney are able to create compelling images that can be used in your marketing efforts. These tools are already having an impact on the traditional stock photo industry. In fact, many people in our industry think that the end is near for the major stock photo services.
Due to the unique images produced by these systems, it is unlikely that Google Search will penalize them for being AI generated. Now, we can expect unlimited creativity from these new systems.
