Fostering AI Adoption

AI is overwhelming.

I think the reason why is some users aren’t sure what to do with it. It’s like the “world’s your oyster” problem. It has so much potential, we don’t know where to start, then when we do use it, we are often underwhelmed by the result.

Many pro’s dig into the weeds of AI, making amazing products, using MCPs, creating RAGs, using agents, attaching it to datasets via APIs; teaching AI how to produce answers, and great showy posts on how they have capitalized on AI. They have the time and bandwidth to tinker and test, and also have the technical, and developmental acumen to create something amazing.

The average user then thinks “oh, it can do so much, let me try”.

Then the average user get disappointed because they are often given a single chat window and told “ask it anything”. Then couldn’t create those results the flashy marketing or technical expert shared, or are told something so wrong they don’t trust it anymore and are completely turned off of it.

The disappointment settles in, and the pro’s who have spent many hours massaging AI prompts to get a great answer say “well you didn’t use it right”.

In the end, yes, the pros will make waves of improvements in their workflow, they will be praised for their understanding of the new technology.

Meanwhile, the rest of us, who are still excited about the technology get relegated to the “what’s the weather like outside” questions, which then may drive development down the wrong path of simple prompts that don’t actually do the things we were promised AI could do for the general user.

My thought?

Don’t tell users “it can do anything”, tell them “here’s HOW it can do anything”.

Users? Don’t get frustrated and quit the first time you use it. Iteration’s take time. Keep picking it up and being curious.

Employers? The focus on AI productivity only happens when you teach CURRENT employees how to use AI. Don’t fire your greatest asset to hire a AI expert in hoping to get fantastic gains from AI. Yes, you will get some of those gains AI provides but you lose the humanity. Humans change, AI still need to be told how to change with those users.

Those assets you just let go before training AI will no longer be able to provide the context needed to continue to make AI work for you. Instead of firing your customer service agent, use the customer service agent to continue to monitor the AI agents asking and answering questions, making small changes to context AI can’t pick up on yet. Then watch as the customer service agent can now handle 5 requests at once perfectly instead of 1. Better yet, watch as they increase customer retention AND handling 5 requests because the user knows there’s a human monitoring and can intervene to just the human the customer needed at that moment.