Artificial Intelligence (AI) is the most discussed technology today, in technology circles, businesses, and personal use. There is little doubt about its short- and long-term impact on just about everything we do. I equate its overall effect on how we live and work to be multiple times greater than the impact of the Internet, and we know how that has changed lives in the past 20 years. For leaders in the rental equipment industry, understanding this shift is becoming increasingly critical.
But what is AI, really? And how can you intelligently harness the power of AI to improve your life and the success of your business? Having seen many new technologies emerge over the past 45 years, one thing has always been true — technology marketers will contort and exaggerate what their products do to exploit the buzz around whatever is a “hot topic.”
That was, and remains, one of my primary concerns for AI — that you pay a premium for some new technology because it says ‘AI’, when it is not. I have read several articles promoting AI capabilities that were clearly not much more than traditional analytics. The primary purpose of this is to help demystify AI and equip you to make informed, intelligent decisions related to how and where you invest in it.
Note: In the interest of full disclosure, I used AI (specifically ChatGPT) to provide information for this article, partly as an experiment and partly to gain insight into specific topics. I found it very helpful in providing coherent information relevant to my prompts. ChatGPT is referenced where I utilize quotes from various questions (prompts).
What is AI?
According to ChatGPT:
At its core, Artificial Intelligence is the simulation of human intelligence in machines. These machines are programmed to mimic cognitive functions such as learning, reasoning, problem-solving, perception, and decision-making. The goal is to create systems that can perform tasks traditionally requiring human intelligence, such as understanding language, recognizing images, and making decisions based on data.
In short, AI can do a lot of thinking for us, and in some cases, act on what it learns. How well AI thinks and what it can do ranges from very specific (called Narrow or Weak AI) to very broad (called General or Strong AI). We utilize Narrow AI in many tasks today — things like voice assistants, recommendations on websites, and image recognition. On the other hand, General AI (more closely resembling human intelligence) is still mostly theoretical and, according to many, a decade or more away.
Types of AI in Use Today
Given that most AI today is Narrow, how is it used? Basically, there are two types of Narrow AI: Generative and Agentic. They can be combined to form Embedded AI.
Generative AI is used to generate content–text (like parts of this article), computer code, images, audio, and more, based on prompts. This is the most used form of AI today (ChatGPT, Copilot, etc.). You ask AI a question, and it researches it and provides you with an answer. The more specific the prompt, the better the answer. Think of this as your personal intern.
In rental equipment operations, this technology can generate customized safety briefings for different types of machinery, create equipment operation manuals in multiple languages, and draft emails to customers about weather-related delivery delays. It can also analyze customer spending patterns to optimize pricing and discounts or prioritize calendars and schedules based on specific criteria.
General AI is still mostly theoretical and, according to many, a decade or more away.
Agentic AI is an agent (robot) that achieves a specific goal, like completing a task or making a decision. Unlike Generative AI, which provides humans with information, Agentic AI takes action based on what it learns. Agentic AI is evolving rapidly and will dramatically improve overall human performance, particularly in business. Think of this as your executive assistant.
For rental companies, this technology can automatically schedule preventive maintenance when equipment reaches certain usage hours or reroute delivery trucks when a customer calls to change their address. It can also predict seasonal utilization swings with your rental equipment and adjust rates accordingly within set parameters.
Embedded AI combines elements of both Generative and Agentic AI but embedded in applications you are using (like Salesforce or Hubspot). When turned on, it monitors what you are doing, provides feedback, and/or takes specific actions along the way.
In rental management software, this technology can pop up alerts when a customer’s rental history suggests they might damage equipment or automatically suggest alternative machinery when their first choice isn’t available.
AI vs. Traditional Analytics
At its core, Generative AI produces useful information from data. It happens that the ‘data’ for AI can include, in addition to your data, any public data on the internet. That’s a vast amount of data. This can blur the lines between different types of business analytics. Because Generative AI produces information based on data, it seems natural to include Business Analytics as a significant subset. The key difference is the data being accessed — your specific ERP data, or data generally found in the cloud.
The more organized and rationalized the data, the more efficient and accurate analytics can be with it.
In a relatively recent newsletter from Forbes, they suggested that everyone understand the difference between types of analytics. Why? Primarily, so that you do not build or buy something promoted as AI when it is basic, readily available analytics. Forbes defines the progression of business analytics as follows (Forbes definitions are set in quotes).
Data Analysis — Providing “data from past events” and presenting results for a human to “review for patterns” and act on.
Predictive Analytics — Looking at data from past events, “making assumptions and testing those assumptions in order to predict future what/ifs” and presenting specific predictions and suggestions for humans to review and action.
Artificial Intelligence — “Machine learning that analyzes data, makes assumptions, learns and provides predictions, all done at a scale and depth of detail impossible for individual human analysts.” The results can either be presented to a human for review and action or fed to an AI Agent to actually take the action.
All three require data, and the more data the better. One question is whose data. Is it just yours or yours combined with other data in the cloud? Regardless, consider these two things.
- The more accurate the data is, the more accurate the results will be. The adage “garbage in, garbage out” still applies, and it doesn’t matter whose data is included. Just because data is from the cloud doesn’t ensure its quality for your request.
- The more organized and rationalized the data, the more efficient and accurate analytics can be with it. Answering some analytics ‘questions’ takes a huge amount of computing power, so making it more efficient is important. But organized data also affects the quality of the result because disorganized data can lead to inconsistent information and/or the ability to connect certain data to the question being asked.
According to Forbes, 85 percent of AI initiatives currently fail. They primarily fail because of data. The data is either too dirty or too unstructured to produce meaningful results with AI.
Let’s break it down.
Data Analysis — Providing data from past events and presenting results for a human to review for patterns and act on. This is generally done via reports or queries that provide information about something specific, such as equipment utilization rates by category, overdue rental returns, or maintenance costs by machine type.
These can be unfiltered lists that you search for what you want to see, filtered/exception lists that show only data falling outside of some specific ‘norm,’ or dashboards that display data similar to an Excel Pivot Table.
Predictive Analytics — Looking at data from past events, making assumptions, and testing those assumptions to predict future what-ifs and presenting specific predictions and suggestions for humans to review and act on.
Consider compact excavators that have been rented heavily over the past three months. With just this information, you might consider purchasing additional units. However, examining historical patterns, demand typically drops by 40 percent in the upcoming quarter due to seasonal construction slowdowns (fictional data, does not represent factual trends). Rather than purchasing additional excavators, postponing that investment for several months would be the smarter financial decision.
Artificial Intelligence — AI goes even further, and can be focused not only on your data, but it can also be broadened to include data from others in the cloud.
Done correctly, broadening the data for analytics can make the result statistically more accurate. AI can also examine far more variables than those typically used for traditional analytics. It is this difference in accuracy that enables you to consider having an AI Agent act on the result.
The only question you need to ask yourself before creating the Agent is, “What is the downside if AI gets it wrong?”
View the original article and our Inspiration here
Leave a Reply