Stop AI Washing: Using AI as an Ingredient to Deliver Measurable Business Outcomes

Every day we hear more and more companies announcing their latest and greatest suite of AI intelligence products, despite not sharing many clear methodologies, or use cases with results. While AI certainly has the ability to transform enterprises in many ways ranging from creativity to productivity, the idea is to use it responsibly to augment man, not replace him. Another central point is to not overpromise on the tech and under deliver on the results just to satisfy investors or get more customers as that is dangerous and what is referred to as AI Washing.

AI Washing Explained

If you don’t know what it is, here’s a primer:

• When a company says it uses AI with no clear explanation of the methodology behind the application, how it was developed and by whom. Real scientists are needed, as is significant time and money spent on R&D—go figure. Black boxes are a no no.
• When AI services include developing AI-powered brands, paired with using random buzzwords like LLMs (large language models) and Generative AI, with no intelligible understanding of the services being provided.
• When a company states it has AI offerings but yet can’t cite what the tools are (beyond random names they assign to  them), explain how they work, or the results they have yielded to date.

All industries are at a crossroads because of the advent of AI, but marketers and insights leaders as top communicators, must establish responsible dialogue and share use cases around what AI really is and is not in order for the tech to do what it’s intended to do which is make man stronger, better and faster, not replace him. This and only this will uncover the future innovation the power of AI and other tech can unlock.

As a former IBM Futurist and AI innovator in the area of agile emotional insights and intelligence, I use it as an ingredient married with data science to precisely score human emotion. This is all done with a specific intentionality around using the best technology that exists to help evolve market research by updating outdated means of customer understanding.

In my experience, I have learned the key to success is ensuring that AI is not used in isolation and always includes direct human input. Critical to the use of AI is also a specific application against a pain point or problem to be solved so tangible results can be measured in order to test efficacy of course, but also to ensure no harm is being done at the cost of a benefit such as let’s say speed or cost efficiency.

Powering Deeper Customer Understanding

My team and I have worked to solve for better means of customer understanding using specific AI applications. We do this via emotional lookalikes that act in essence as anonymized customer experience “test dummies” that allow the use of math and data science to put guardrails around AI. Why? Because in doing so, myriad unconscious biases that often result from AI used without direct human feedback, are removed, at the earliest stage possible in building the customer journey, to maximize impact.

A great example of our work was highlighted when we were hired by Derivative Path, a SaaS tool designed for mid-sized banks and credit unions to do their own hedging. Their challenge was understanding what their buyers who were corporate treasurers and investment fund managers would need to emotionally trigger them to try such a service. What our data uncovered was that while the ability to do hedging on their own was empowering, there was also a lot of fear in unknown tech and taking on more responsibility.

Our emotional AI enabled us to uncover these challenges and helped the company create a human-led go to market strategy that helped invert their sales and messaging approach by leading with the human led-advisory services they offer, versus the tech alone. According to Derivative Path’s head of marketing Terrence Richardson, our responsible use of AI and data science was “transformational and set the company up for better business success and improved business outcomes.”

Leveraging Emotional Data to Evolve Accuracy of Consumer Sentiment and Combat Misinformation

Another responsible use case of AI married with data science to drive improve intelligence that led to better outcomes was the launch of an Entrepreneur Tracker, with our client Clutch, a boutique consulting firm designed by entrepreneurs for entrepreneurs. With them we set out to take the emotional temperature of entrepreneurs against discretionary and non-discretionary spending, but also against things such as how they felt about being an entrepreneur, the state of their business, and their overall sense of opportunity for successful entrepreneurship, in today’s environment.

The results were astoundingly compelling, as while the media had used traditional spending data trained by various AI models in isolation to paint quite a bleak picture of the landscape, our modeling which included once again marrying AI with data science and direct human input, uncovered a very different picture, namely that entrepreneurs across four generations feel we have entered a global age of entrepreneurship where there is enormous opportunity for both personal growth, but also the opportunity to help make the world a better place.

This example is powerful in a different way, not just in highlighting key differences in the data to improve a company’s GTM strategy, but also in dispelling the type of misinformation that can be generated by technologies like AI when used as a catch all to supercharge data, without human input.

AI Washing Must End

As 2024 unfolds, AI continues to be the bright shiny object that everyone is buzzing about, yet few really know what it is beyond a surface level. There is even widespread discussion of how AI will suddenly make humans obsolete in every industry from technology to journalism.

However, what too many people are overlooking is the slippery slope that exists when AI is used in isolation without any type of ongoing direct human input. We are pretty certain that while it is amazing that one can now use ChatGPT or Generative AI to write a cute lullaby for their child, or enable a journalist to generate potential headlines in a nanosecond, there are circumstances where in the wrong hands, this ability to artificially generate human dialogue, left unchecked, can and will lead to potentially dangerous outcomes.

On the flip-side , uses of AI that are responsible and actionable can and will lead to the next generation of business growth and the only way to ensure that the latest trend of AI Washing remains just that — a trend, versus a systemic business and societal concern.