A Key Misconception About Generative AI
Generative AI has exploded in sales and marketing as the panacea for all our content woes. Need a subject line? An article? A product description? Just ask the platform, and it generates one for you. There’s an underlying misconception about this GenAI, though.
Let’s back up for a second. At its root, AI is about systems that can store and analyze data to predict what will happen next based on what happened prior. If you can feed terabytes of consumer activity into AI, it can likely predict what a consumer will do next based on their past behavior. If you’re a sales leader, you can likely forecast your sales. That type of information is priceless to a company. The more data you give these systems, the more accurate the outcomes.
GenAI doesn’t work any differently. Words are numbers. As AI analyzes billions of words strung together, GenAI predicts what will likely come next. It’s a forecast of what is statistically probable. As a result, many assume GenAI is creative or remarkable in its responses.
Sales and marketing are both art and science. While a sales professional can use AI to prioritize their next activity, the art is executing it. The same goes for marketing. AI may reveal that a customer is more likely to take action with the right personalization and segmentation; a remarkable offer or marketing message will drive conversion.
As humans, we have a distinct advantage over GenAI (currently). We have a lifetime of experience, multiple senses, the ability to process input from all of them, and the ability to interpret and immediately leverage that data. This doesn’t mean that a GenAI response isn’t valuable. I utilize GenAI throughout my workday. It helps me research subjects, summarize high volumes of content, and validate my understanding. GenAI is my coach, my research assistant, and my grammar and spelling checker. GenAI helps me to merge science with my art to produce exceptional results.
A sales professional may visit a client, see an autographed football on the shelf, and connect with the prospect over being fans of the same team. A marketer may discover a unique opening act for their favorite band, find an audience that matches their target persona, and license that music for their next campaign. The source data and models that trained GenAI have yet to train on this data (for now).
Companies that want to be exceptional won’t get there with statistically probable content and strategies. That’s not to say GenAI can’t help. Statistically probable content likely will not drive exceptional response. That doesn’t mean it’s not valuable, though. It may very well perform better than content that’s produced without. It may.
With every company jumping on the GenAI bandwagon, I have little doubt that search engines are being inundated with billions of AI-generated pages today. Those companies may grow their market share by using content-production efficiencies tailored to their target audience. Will they see exceptional growth? Doubtful.
The irony of AI-generated content is that this statistically probable content will then be ingested to train the next statistically probable GenAI outputs. Without feedback on whether that content is producing real revenue for companies, we may be heading down a road where probable content becomes the lowest common denominator. It may simply lead us to boring, unremarkable content.
The key to GenAI’s success isn’t its ability to spit out statistically probable content. It’s to connect those systems to your business activities to generate content optimized for your brand, culture, and customer personas. At this point, you can have optimized GenAI content which will, no doubt, produce better results.
Ultimately, as the datasets that train these models merge with the real world (vision, sound, touch, smell, and taste), perhaps we can inject human creativity too and see truly innovative results from GenAI. Will we ever get there? I believe there’s a statistical probability it will happen in our lifetime.
This article was originally published on Forbes.
