Avoid garbage in, garbage out trap of AI business tools

Opinion

There are many articles, perspectives, advice and cautions regarding the use of artificial intelligence appearing daily across all sources. The big question remains: does it work?

The short answer: “It depends.”

There is no “EASY” button that will solve all your problems. Just because you outsourced some business activity to an AI tool doesn’t mean you are automatically going to be more productive or provide more value to your customers.

The long answer (also): “It depends.”

If your aim is to have an AI tool create something special in your business, the overriding question is, “Does it add value for your customer?”

You must ensure the output from your efforts is transformative in some way. This is what a tool should do for you — make something easier or better. However, your output is a direct correlation to your input. You cannot default your thinking or use of the tool to be an easy solution. Your results may not be as special as you hope.

Remember the first time you used Google to conduct an internet search? Often, the searches did not deliver on what you really wanted. So, you refined your parameters and choice of words to find what you were looking for.

The same consideration holds for any AI use. While it is fun to test and trial different prompts and AI tools (I strongly encourage this approach), there is still an important human element that brings your output to life. What is the real reason you want the tool to help you do?

So how best to use an AI tool to craft improved inputs to avoid a “garbage out” response?

Faron Hiebert, chief delivery officer at Encore Business Systems, describes how the firm positions AI for clients.

“It’s not just about adopting the latest technology, but about integrating it in a way that aligns with our overall business goals” he says. “By enabling AI for our customers, we empower them to leverage these advanced tools to drive their own innovation and success.”

According to Hiebert, the key point about adopting AI is you must align with business goals. (This perspective reinforces my points that you need to know what your objective is and not just think AI will solve everything.)

This approach can be described with the following three steps to avoid creating a “garbage in” situation that ultimately creates poor results.

First, determine what your business bottlenecks are or where you want to add more value for customers.

This requires thinking first and not just selecting a tool and diving into submitting non-specific prompts. If you do not have a business plan, how will you know what AI might be able to help with? Aligning it with your business goals first is a critical success factor.

Second, ensure you can clearly describe what you want the tool to help with — what is your “win?”

This is the time where the wording for the prompt and the type of answer you are hoping for are determined. For example, is this where you need some assistance in generating ideas for a new product? Generating ideas with AI tools can be helpful, but the results are based on combining only existing data from the web. Do not expect any new options. You will need to review the ideas and combine and adjust them to find something that demonstrates value for your customer and that you are capable of delivering profitably.

Third, do not accept the results as final.

I am willing to suggest, based on experience and discussions with my clients, that there are so many additional thoughts that will enter your evaluation process once you see the first answer the AI tool provides. Consider using a “yes, and …” approach to ensure you have all the right considerations included in your search for an answer that can be implemented.

As mentioned earlier, if you do not use a specific and well-thought-out prompt, please do not expect a well-reasoned answer. Take your time to evaluate what you want and how this fits into your business before thinking perfection is only a mouse click away. Business success is never that easy.

Tim’s bits: There is something personally satisfying when you solve a problem or produce an intriguing idea. I am concerned too many people are not willing to put in the effort to achieve this satisfaction. If we outsource this type of joy, it actually feels like cheating. And this is not a long-term strategy for success because no winning game plan is built on cheating.

Tim Kist is a certified management consultant, authorized by law,

and a Fellow of the Institute of Certified Management Consultants

of Manitoba

tim@tk3consulting.ca