Find Your Sweet Spot with AI

When faced with a new technology like AI, people and businesses go through a predictable learning curve. This happened back at the birth of the internet, the launch of ecommerce, process automation and now with generative AI. As these cycles repeat, some people and companies move faster through stages of understanding the technology, trying it and then putting it to productive use for themselves.

This evolution of learning and applications is distinctively different for individuals and for organizations but there is a lesson in understanding both - as it opens career opportunities. 

We have called these evolution cycles 4E. Let's take a closer look.

Individual AI learning curve

To make sense of AI we go through 4 stages:

Personal AI Stage 1) Entertainment

In this stage we are trying out the new technology, playing with it. Trying to make sense of it for various situations in our lives and daily workflow? Can I create a birthday card with DALLE? Can I complete my homework for school? Can it help me think through a report I have to write for a meeting? The initial results in this stage may not exceed the quality from our current results, but we also get a glimpse of what is possible. For some of us, this stage is sufficient and we can turn to this tool occasionally without changing our lives and workflows. Some of us want to dig deeper and understand how to use the tool better to produce more predictable results. This leads to the next stage:

Personal AI Stage 2) Education.

We dive in and learn about prompt engineering. Discover chat forums that share tips and tricks. We may go deeper and learn the underlying language and algorithms. We go to the GPT playground and play with temperature settings. Others go deeper still and try their luck at python code with GPT or Claude APIs. The education phase focuses us on understanding the levers and parameters that we can control to get better results for our needs from GenAI tools. And as we do, we start improving our productivity. We create meeting summaries that generate email follow ups to attendees. The meeting action items will be linked in our todo app. We find a way to scan all birthdays of our contacts and generate and send birthday cards for them. This moves us in the next stage:

Personal AI Stage 3) Efficiency

For those of us in this stage, life becomes an ongoing discovery. We are looking for ways to apply our new superpower to everyday tasks we deal with. This stage is the birth of the AI First Mindset (a topic of a different post).

We apply GenAI to our workflow and we are saving time, reducing stress, and becoming more productive. In this stage our efficiency may or may not be visible to others. We get that report done in half the time, but my boss may not know that. We do a comprehensive meeting summary but no one knows it took us 10 minutes and not 4 hours. Effectiveness saves us time but what we do with that time makes the difference. If the hours saved are used to learn a new skill or degree, then it improves our outcomes, pay and lifestyle. If the extra hours allow us to get more work done, see more customers, process more orders then we increase the company's business output, and this is when we reach the final stage:

Personal AI Stage 4) Effectiveness

In this stage, we focus on moving the needle for ourselves or our employers. This creates better business outcomes, we exceed our targets, improve our OKRs, earn higher bonuses and commissions.

Noone is operating at a single stage all the time. With the rapid change of generative AI and the new capabilities and tools introduced daily, we can keep repeating this cycle. You may be in stage 3/4 using prompts for Claude and GPT4 but your ability to create consistent characters in DALLE is at stage 2. And you are still playing with tect to voice or text to video AI in stage 1. Some will advance most of these skills to stage 3/4 while others have no need for it. It is highly individual, although it is our belief that those operating at stage 3 or higher will outperform others staying at lower levels by a significant margin. And this outperformance will show up in better compensation, job opportunities and overall life outcomes, whether you are in business, education, healthcare or in creative arts.

Our poll on Linkedin showed where respondents were in April 2024

While organizations are the sum of individuals in these 4 stages, there is a higher level of GenAI evolution and maturity in a business.

Companies also go through distinct stages of AI maturity.

Both people and companies arrive at the same destination through slightly different paths.

Business AI Stage 1 - Experimentation

the companies start various GenAI projects. Budgets are made available for these experimentations. There is a more discretionary nature of funding without the usual rigor of justification. Leadership teams think they have to react to a technology disruption. They bring in consultants to learn and try the new technology. This stage is the realm of proof of concepts and use cases. The business does not consider the tools proven or "enterprise ready" just yet. Some of these experiments will show promise and actually create productivity gains, better customer interactions, faster processing time.

Business AI Stage 2 - Enablement

The company will train employees on the tool, roll it out, and encourage usage at a broader scale. Executives will monitor business impact. This stage will result in hours saved, work quality improvements, and improved customer satisfaction. The results may not be consistent and repeatable. The business will keep tweaking the knobs, refining the processes.

Business AI Stage 3 - Execution

Operating with the new tool is now standard operating procedure, how we do things. The usage of the tools is consistent but may not improve business outcomes. A good example from recent years is collaboration tools like Slack and Teams. There is consistent usage, a general sense that it helps with the business but no expectations of measurable outcomes. Are there fewer meetings and emails because of these tools? Do the savings on meeting and email time exceed the cost of these systems? Most companies don't know. This is part of the process. Part of efficient execution in the business. The typical metric in this stage is hours saved improvement in documents processed, customer tickets handled and order processed. 

Business AI Stage 4 - Effectiveness

Once these tools start impacting the basic metrics of the business like revenues, profitability, products shipped, etc, the business is in the effectiveness stage. Many teams struggle to distinguish between efficiency and effectiveness. Most companies lack the value creation discipline to measure impact of projects, especially outside the core business, like in finance, HR or IT. At its simplest level, if you are running processes with less manual effort in less time saving hours of work, you are efficient. Once these process efficiencies translate into consistent reduction of cycle time, costs, increased revenues, higher production yield, better customer service, then you are becoming effective. The line is often blurred between good execution (efficient operation) and high effectiveness (increased output, profit, revenues). Often the problem lies in the fact that we’re automating processes with AI that do not move the needle, so to speak. Process and task selection for AI transformation is critical for the business to gain the benefits of AI transformation. In the AI enabled effectiveness stage we should see clear improvement in metrics linked to revenue per employee, improved cost of revenue, higher production yield, shorter first resolution time in customer service, higher sales conversion, etc.

Shaping your career: Matching your individual AI progression to the company’s AI maturity

At any point in time we and our companies can be in a different stage of AI maturity. Some people can be ahead of where their companies are. While others may be falling behind in their mandatory AI training. Both of those can be detrimental to our careers, job satisfaction and opportunities.

We often think of Mihaly Csikszentmihalyi’s famous research on Flow states. For you to be in “AI flow state”, your employer’s AI maturity should be in sync with yours. If you are a data science genius who rolls her LLMs on Hugging Face and your company’s leadership is struggling to find AI use beyond meeting summaries, you will be bored, frustrated and probably pursue your AI interests outside the business.

On the other hand, if you are just learning about AI and you see the company promoting advanced AI use in every function, you may be stressed from trying to keep up. 

This is true even beyond AI. Ideally, we all should work somewhere where the people around us are of similar maturity on technology, processes and work standards. That is a work ‘flow’ state. You are challenged to stay a bit ahead of the flock but not stressed out for fear of falling behind. Or you are not so far ahead that you are bored senseless where people are just “not getting it”.

For the company’s leadership team, if they want to be an AI leader, they should have employees interested in advanced AI maturity. That creates a flywheel effect. Being great at AI is not everything. But if AI leadership is our personal or business mission, we constantly need to advance and mature our AI mindset both as a leader and a business.