Every board of directors couldn’t be blamed if they were now pressuring their executive leadership to explain their company’s AI strategy. Not only for their own products and services, but also for how to will deploy AI to optimize workflows and reduce unnecessary human interactions. On the surface, these are the right questions. It’s what an organization chooses for answers that might cause some issues. Let’s talk through it.

The Interface Part of AI

There’s a continuum when it comes to what a customer/user/client wants. It’s fair to say the best of all worlds is “do it for me.” I go to the restaurant and you feed me. The next notch is “do it with me.” I order one of those delivery services with chopped up and measured meals. After that is “do it yourself.” Go to the store and figure it all out. If you want, I’ll give you a recipe, but that’s it.

For any company building technology right now, at this point, I think at least an AI interface is a must. We’re going to shift so quickly from “Oh, how clever” to “where the heck is it and why isn’t it smarter” quickly. I already am. For my new gig, I’m using Microsoft Outlook. I clicked on Copilot (the MSFT version of GPT) and asked it to filter my inbox. It said, I’m not doing that, but I’ll tell you the steps if you want. (grrrr.)

If you don’t make it easy for the AI to help me use your app better, you’re on a ticking clock, in a lot of cases. And I think this spreads to most any app, big or small, out there, except tiny apps.

The Bigger Side of AI – Agents and MCP

Not all that many years ago, we were just getting into the idea that software needed an API, so that if the company didn’t want to build out every little feature, they could just produce an API (application programming interface) and let other people write code that talked back and forth to the API.

In the AI world, agents (a bit more than your typical LLM) can talk to a company’s MCP (model context protocol). In AI, the MCP is an open standard that provides a standardized way for AI models, especially large language models (LLMs), to connect and interact with external data sources, tools, and services. It’s how people will be able to ask their agents things like, “Can you book a trip for me to go to Vegas to see my Dad on Thursday,” and have it actually know most of what needs doing.

If you’re building websites that you want to be worth a damn, you’re going to have to contemplate the MCP stuff and how to make it all work. The MCP will be the kind of guardrails that will make your site more useful/valuable to our new robot friends, and if you are in competition with another company and their site has an MCP and yours doesn’t, well.. that might be all she wrote.

Your Team Must Use AI Tools Where Prudent

If you asked your finance team to crack open the spreadsheet showing next quarter’s shortfall that you’re tracking, and they produced sheets of paper, pencils, and rulers to keep the columns straight, what would you think in that moment? If you see your engineers working feverishly on their abacuses (Oh wow – I just had to look up the plural to be sure), there’s a problem, right?

The sheer number of applications to where AI accelerates tasks and augments many teams’ ability to produce more (and better) material faster is such that it would be imprudent (and maybe negligent) to overlook the application of AI helpers.

Let’s say I ask you to produce every trouble ticket generated in the last quarter and group them by problem type, but only show me ones where the customer reported a problem known to be caused by the cloud version of the code and not the data center. If you’re not asking Rovo to do that for you in Jira, you’re wasting everyone’s time.

AI summaries should come at the top of any report more than a few pages long. Heck, reports should be fed straight into your GPT so that you can query the information ahead of doing any deeper thinking. (Yes, thinking by yourself is still quite important, even with fast AI tools.)

There are very few parts of any modern business where an employee shouldn’t be using AI as one of their daily tools to facilitate better work turnaround. (Yes, this will require some training and coaching to make it better.)

2026 is a Disruption Year

I believe we’ll see some incumbent business leaders topple in the face of more nimble AI-augmented organizations that originally had no chance in hell of beating their larger competitor. I think it will come in the form of rapid iteration, of augmented intelligence and data operation, and in building more visible and accessible workflows with companies outside of the primary organization.

Between building agents and the power of MCP opportunities, I believe we’ll see some real changes that will require rapid reaction for some organizations to compete against.

I don’t say this to worry executives. This is my attempt to give you more to think about in your planning, to give your operators more to consider in how they look at their current operations plans. I believe a lot of change is coming fast in 2026.

Are you ready?

You know I’m here to help.