We presented the following at an event for a Third South Solutions client in July.

Whether we work for our own companies or for clients, it often takes time to move up the curve of AI implementation. A chart recently published by Notion illustrates four kinds of AI maturity:

Notion defines each of these categories: Thought Partners help individuals with isolated tasks (e.g. write an email that says…). Assistants help individuals with the added bonus of being context-aware (e.g. take notes on this customer call). Teammates work collaboratively with employees, deploy agents, and own a scoped piece of a workflow. Systems run critical workflows around the clock and use recursive improvement to get better with more usage. 

Reducing the four categories into two buckets reveals an extreme distribution: 88% of businesses are in the first two categories, just wading into AI. Only 12% are using AI in a way that creates larger business improvement. The difference isn’t from a lack of trying, but instead what they’re pointing it at.

The 88% are mostly people using AI individually to solve their particular tasks. They write better emails, clean up code, start the draft of a presentation, and summarize meetings. There are certainly time savings from this type of AI use, but it is often isolated and negligible. AI can help one person be a little faster at their own job through individual experimentation, but there are more exciting uses for such a powerful technology outside of using it to build better SparkNotes.

The 12% point AI at the business itself. It’s wired into their infrastructure. Someone owns it end-to-end and it generates real value for the organization. For example, we’re working on a new project for a client that helps the account management and engineering teams generate new sales proposals. By referencing prior sales proposals and accessing data across multiple internal systems, the team will be able to deliver more complex proposals more easily, shrinking delays for customers. This project will also deliver incredibly rich context about the client’s internal processes. That context will serve as an important foundation for future projects.

Perpetually staying in the 88% leaves opportunities on the table for teams. Saving fifteen minutes a day on email drafts is progress. But true leverage happens when AI moves from a niche productivity tool to an asset that can create change in your organization that goes beyond the individual.

The jump from 88% to 12% takes three things:

1. Buy-in from the organization to deploy AI tools, including ways to monitor success and mistakes

2. A culture willing to tackle cross-functional workflows with AI

3. Technical execution to deploy AI in a thoughtful, effective, and secure way

To get started on making that jump, identify a small, routine, low-risk task that when solved will clearly demonstrate value with a quick win. Bring your team along to show that the goal with the project is fundamentally to help them and the organization run better. Build the tool so that it functions effectively and securely, while keeping people front and center to check and improve the outcomes. Do that enough times, and going from the 88% to the 12% starts to feel like a hop instead of a giant leap.

The full presentation is attached to this email. Give it a look and reply with any thoughts you have.

Best,

Myles and the Third South Team