AI Is Not a Magic Wand For Your Business

*A Jasper AI generated image of a robot holding a magic wand.*

Article by Chris Miller

AI is part of the daily conversation, but for business owners, I think we are all struggling with what embracing AI for our business looks like. I’m not talking about the growing number of off-the-shelf apps and tools which are available, I mean solving the unique problems each of us face and the overwhelming prospect of how AI applies to those specific needs.

AI is all but touted as a magic wand, but it’s not that simple. One of the most frustrating things for me is trying to find the use cases that add tangible and measurable value to a business. Gee-whiz novel solutions aren’t going to move the needle.

Right now thought leaders are talking about use cases, much around marketing which include content generation, personalization, text summarization, etc. The problem is - these aren’t “use cases” they are broad applications of the technology. A use case to me is “Client A solved <painful problem> by deploying AI <in this way> which resulted in <measurable metric>”

As an AI practitioner I’ve spent a lot of time participating in the ecosystem, especially in the Year of the Large Language Model. What I came to realize is that the reason we aren’t hearing about a lot of real use cases is because we collectively haven’t gotten there yet. You Are Here, and it’s the Wild West.

Flushing these use cases out today is going to start with real conversations which explore these business problems. Not every business problem is an “AI problem”, and some use cases AI can solve may not provide enough value or fast ROI.

There is no doubt early adopters will gain advantage by realizing improvements to business problems, whether that’s employee productivity, improved customer support, or process automation. Those early adopters will not be risk-averse, they are curious and explorative, and have the vision for solving the hard problems now so they can focus on other things in the business.

Chatbots are the current buzzword, one I don’t prefer because it makes me think of being held at arm's length from a human on a website. It’s true, a lot of power in AI can be derived from “conversational bots”, and we have tools to build those (amazing) experiences right now for anyone.

What is often skipped in the discussion is known as “LLM-enabled applications” in which the LLM is used as a reasoning engine to analyze a problem, explore solutions, and lay out steps to solve. This can then be applied to a number of domains. This is an area I feel is ripe with opportunity and represents the ability to automate processes and truly augment our daily work lives in a positive and productive way.