AI explained: an introductory guide to keeping up with Artificial Intelligence

Artificial intelligence (AI) tells a story about technology’s evolution—productivity, innovation, and the relentless push forward. It’s difficult to measure its influence when we are in the midst of adopting it (how could we know how easy spell check, online shopping, or toll booths would become, in hindsight?). As we jump into partnerships with generative AI and apply it our daily tasks, we’ll have to be aware of and monitor developments in its weaker areas such as transparency, accuracy, and bias. 

To unpack the potential impact of generative AI tools, it’s important to start from the very beginning: what is AI? From what and how did it evolve? And how can it support Customer Success professionals in their day-to-day jobs?

The beginning: AI’s origin story

Artificial intelligence didn’t spring up suddenly in the last few months (despite how it feels with the tidal wave of attention over OpenAI and Chat-GPT). The process was iterative and in the works for decades. Here’s a snapshot overview of three common types of AI and their development.

Machine learning

Quick definition: learns what you teach it, studies it, then tries to operate on its own

• Gives computers the ability to learn without being explicitly “told” (i.e., programmed)
• Uses statistical techniques to learn patterns in data that can be applied to a variety of problems (classification, clustering, regression, etc.)
• Has a wide range of applications, from fraud detection and medical diagnoses to self-driving cars and voice recognition software

Natural language processing

Quick definition: processes text to tell you if it’s positive or negative

• Is a subset of machine learning that reviews large datasets related to language
• Employs a rules-based methodology. Tags different components of the data (i.e., subjects, verbs, objects) to establish connections and determine intent or sentiment
• Can be trained to recognize positive/negative sentiment similar to humans using machine learning algorithm on a data set (example: published articles)

Generative AI

Quick definition: analyzes tons of data with help from deep learning to do more than just spit out summaries

• If machine learning is the parent, generative AI is the high-achieving child that also got a scholarship from the school of neural networks and deep learning
• Analyzes larger sets of data with unique outputs (AKA new data that is similar in some way) and can cluster information to create probable next steps
• Aids in a variety of tasks: image generation and recognition, text creation and editing, voice synthesis (example: Chat-GPT)

The dilemma: AI’s evolution and what’s next

From Clippy the Microsoft Word assistant to recent leaps with GPT-4, we’ve seen a fascinating evolution of helpful tech tools for our daily tasks and jobs. Things like spell check, auto-correct, and now suggested replies and auto-generated text are becoming normal practices in how we work.

In fact, the global AI market is expected to reach $110.8 billion by 2030, according to research from Acumen research and consulting.

While we continue to integrate tools like Chat-GPT or DALL-E (an impressive image generation tool) at a rapid pace, it’s important we keep our humanity in mind. Anytime you use a tool like this, pause to ask yourself:

• Was the tool’s response helpful this time?
• If it wasn’t, how I can better get the information I need? Or, is this the tool I should be using right now?

Consider these points as you integrate and promote generative AI tools into your own workflows and software. Use them to spawn further discussion around the benefits and setbacks of adopting such tools.

The potential: how to use AI right now

How can you take advantage of easy-to-use AI tools right now? While AI is not sophisticated enough to fully take over your decision making and problem solving, it can digest data to help you make decisions and generate content faster.

Here are some helpful use cases:

• Easy customization (example: write an email)
• Research (example: tell me about a brand/company, specifically their business model and PR/marketing strategy)
• Better communication (example: rewrite this email for clarity.)
• Ideation (example: suggest strategies for how restaurant owners can use more press releases)
• Content assistant (example: summarize this transcript from a video meeting and keep word limit to X)
• New products (example: write a press release using the following details to promote product’s enhancements)

How a software platform can use it

We’re already reaping the benefits of using AI at ChurnZero. We started with our Success Insights (which uses machine learning algorithms) and took usage data and customers who have churned or renewed to identify patterns that would put them into at-risk categories. We found the biggest risk to be login rates in the last 30 days, license utilization in the first 90 days, and the percentage of users using a sticky feature within the product. Now that’s specific intel.

With our newly released Customer Success AI, ChurnZero users can access generative AI directly from our product (bonus: it’s open to anyone who’s interested in trying out generative AI, customer or not).

Customer Success impact

Given that Chat-GPT broke the record for fastest growing user base for an app, it’s no surprise Customer Success has jumped on the productivity bandwagon.

For Customer Success teams, the impact is similar to other automated systems—there’s less tedium and more time for strategy. What were considered “chores” of the job, like responding to questions and drafting communications, can now be supported or completed by AI.

This doesn’t make strategic work easier—if anything, that component will be more challenging given higher expectations and growing new business. But as generative AI continues to improve with updated interfaces and built-in prompts, the average Customer Success manager will have more information at their fingertips, and more tasks completed at the end of the day.

AI resources to dig deeper

It’s easy to become romantic about new technology. Let’s remember to use generative AI for what it is: a helpful tool with room for improvement. It hasn’t solved all our business problems or developed new thoughts (yet), but there’s no perceived downside to asking AI questions right now.

To learn more, use these resources below, ranging from entertaining to robust:

• The Turing Test – Quick animated summary from Ted-Ed about how Alan Turing paved the way for how we measure a computer’s “human” intelligence.
• The Generative AI Application Landscape – A visual representation of the AI tools currently on the market for individual users (including text, video, image, coding, speech, 3D and more). The corporate landscape is pretty crowded as well, as seen in this recap of the top 100 AI companies in 2023.
• Customer Success AI – Play with ChurnZero’s new generative AI tool for crafting content, brainstorming, and ideating on demand.
• Artificial Intelligence Index Report 2022 – How is commercial deployment affecting AI research and progress? Stanford tracks AI’s evolution in their extensive annual report.
• The State of AI – If you’ve graduated from mainstream articles but aren’t quite ready to read the full Stanford report, then take a spin through this annual slide deck on the general AI landscape.