What do you think of when you hear “AI”? 

Perhaps your mind jumps right to leading “frontier” applications like ChatGPT, Gemini, or Claude. Or, maybe it jumps to the capabilities of automation and eliminating repetitive tasks. You might also think about potential business impacts around efficiencies, accelerating processes, and improving product quality. You might even think about the risks or governance needs surrounding the technology.

Much of what business leaders are considering when it comes to AI tends to be specific to generative AI, which focuses on newer methods that create something new. In reality, artificial intelligence has been around in some form for decades—and with many of these capabilities still bringing incredible business value to organizations. 

As managers, committees, and leadership teams push to integrate AI into their processes, it’s important to zoom out and carefully consider the full range of capabilities that artificial intelligence has to offer. From clearly scripted rules to tapping into creative capabilities, understanding the categories helps you carefully consider your organization’s or team’s unique business needs and what kinds of AI solutions might match to deliver true value. 

The Spectrum of Artificial Intelligence

Even though generative AI is in the limelight (with some vendors taking advantage of the “AI” buzzword to label automation as AI features), think of AI as a spectrum. I find it helpful to think of AI as an umbrella term with a sliding scale that you can use to determine how your own AI implementations should work: deterministic, predictive, and generative.

Deterministic AI: Deterministic AI has explicit, fixed rules that people (usually business analysts, developers, or machine learning engineers) program to accomplish specific tasks. Deterministic AI relies on these predefined rules and algorithms to reliably and consistently perform a predictable task. For example, a deterministic AI system might flag multiple login attempts as potential fraud if the login location is unusual and from an unrecognized device, or it might trigger an outage alert if power usage drops below a certain threshold across multiple meters in the same area.

Predictive AI: If you were a technologist in the late ‘2010s, you probably heard a lot about “machine learning” and “big data” analytics. Predictive AI is heavily related to those technologies. These systems use technologies like machine learning to look at historical data to identify and learn from patterns in order to make predictions about the future. For example, predictive AI might analyze historical trends and market conditions to recommend budget planning for the next quarter, or it might forecast equipment failure by analyzing sensor data and spotting patterns that occur so technicians can replace parts before they fail. 

Generative AI: Generative AI, which has become the “default” understanding of what AI is for laypeople, creates something new like text, audio, video, images, and code for humans. It automates tasks, surfaces insights, and engages users because it takes advantage of a strong knowledge base and a level of randomness. If we continue the example from our deterministic section about triggering an outage alert, a generative AI solution might help craft an outage notification that adapts wording based on the severity and expected duration of downtime. Or, it might simulate customer demand based on different scenarios to help business leaders forecast impact.

Bridging Deterministic, Predictive, and Generative AI to Create Hybrid Solutions

Although the hype of artificial intelligence feels very focused on generative AI capabilities right now, the reality is that many toolsets and systems take advantage of all three types. What might be advertised as “AI” could be a smart set of deterministic rules, predictive decision-making levers, full creativity, or a hybrid of all three. In fact, a majority of embedded AI systems are hybrid combinations of different capabilities depending on the type of experience you want to create for users.

Consider this example: organizations in a variety of industries—from banking and e-commerce to healthcare and utilities—have some form of chatbot on their website for customers to interact with. These chatbots answer frequently asked questions, direct users to specific places on the website, and route users to agents when necessary. In a hybrid system, the chatbot may implement:

• Deterministic AI to provide very scripted answers for sensitive types of questions, such as how to fill out a specific form such as a symptom checker to determine if a situation requires further medical support.

• Predictive AI to flag chatbot interactions likely to need a human agent based on past patterns, such as escalating suspicious charges to a fraud specialist or pinpointing that a user needs troubleshooting help and routes them to customer support.
• Generative AI to answer common or simple questions in a natural and human way based on additional context, such as when a user asks if a business is open, the chatbot can answer specific to the day and time that the user asks, rather than providing all open hours.

Of course, each type of AI comes with its own set of unique pros and cons. If you’re embedding AI capabilities or building an AI system to support your workflows and processes, it’s critical to not only consider your unique use case but also the types of interactions you want to enable for users. Whether you’re looking for a chatbot, an agent, or an autonomous system, you will want to consider how to take advantage of the benefits that deterministic, predictive, and generative capabilities bring to the table. The right partner will collaborate with you to uncover your true business needs and build a trustworthy, accessible, and value-driven solution that balances the best of all worlds.

If you’re interested in learning more about the spectrum of AI’s capabilities and tradeoffs, realistic user experiences, and how to uncover your readiness for an AI project, reach out to us about our workshop AI & Appetizers. Our 90-minute workshop gives business leaders AI clarity separated from AI hype to build alignment within your organization.

Learn More