Should You Take the Leap: A Guide for Effectively Evaluating AI for Your Business

by Ryan LaPensee

‍

This year has seen an absolute boom in the technology space when it comes to Artificial Intelligence. Every day, a new discovery or use case emerges, and is now likely a deluge of information coming to business owners in their email, social media, and news sources. Should I consider using AI? Am I going to fall behind competitors who do? How does it even work? It can be overwhelming when you try to navigate all the information out there in a way that applies to your business and industry.  

As someone actively working in the field of generative AI, I want to spend a little time here helping my fellow business owners better understand how to make sense of AI and highlight how you can effectively evaluate bringing AI into your business.  

Easy First Step

As an easy first step in evaluating AI, log onto ChatGPT and become familiar with this open-source generative AI tool.  I recommend paying for ChatGPT Premium (it’s about $20/month and can be cancelled at any time), but you can also start with the free version as well. When you first log into ChatGPT, give it an easy exercise that will help you see what AI in general is capable of: ask it to craft an email message. You might have an email already written but don’t feel confident about the tone or the structure. Ask the AI to potentially reorganize the email body or have it craft the email in a more professional tone. What generative AI is good at is producing new thoughts from the inputs it’s given. I highly recommend asking it to do this several times and change the wording on how you ask the AI. You will start to see patterns and limitations as you test it out. You will start to understand what the limitations are with ChatGPT as it doesn’t have context of your proprietary data and it’s not interfacing with your internal applications/systems.  

Keep in Mind What AI Can Do vs. What It Can’t

Despite all the hype (and snake oil) to the contrary, AI is not magic. Powerful? Yes. All-Knowing? No.  As you think about how AI can help your business, keep in mind a few fundamental facts about AI as it exists right now (sure, these will change as AI evolves, but for the immediate future, these are facts as we know them).  

What AI Can Do  

• Shorten complex, long, “boring” business processes by 95%
• Provide a rough draft so you don’t start with a blank slate  
• Incorporate multiple contexts at once  

What AI Can’t Do  

• Create brand new content like a human  
• Brainstorm like a human
• Ideate like a human  
• Become any smarter than the humans it interacts with  
• Integrate easily into your existing business systems

Exploring AI Use Cases

Once you’ve gotten a grasp of AI’s capabilities, spend time with your team identifying areas where AI could augment daily operations (in the technology realm, we call these “use cases”). And the more detail the better. Like any technology, AI performs best when it’s applied in a well-thought-out way. Streamlining manual data entry or alleviating communication bandwidth bottlenecks might be potential scenarios where AI could be beneficial. Below are a few examples of how AI could help areas of your businesses:  

Administrative

• Automated Email Responses  
• Document Preparation  
• Data Entry

Human Resources

• Resume Screening
• Employee Onboarding  
• Employee Survey Analysis

Sales

• Lead Generation
• Customer Follow-ups
• Data Analysis  
• Personalized Product Recommendations
• Customer Feedback Collection

Marketing

• Content Creation  
• SEO Analysis  
• Competitive Analysis

‍Critical Considerations

As you would expect, integrating AI into your business has more than a few implications that should be taken into consideration.  Here are what I consider to be key items to consider before making a move into AI:

• Proprietary Data: Exercise caution in sharing sensitive business information with open (not behind your firewall) AI systems. As with all technologies, bad actors exist who will take advantage of AI’s vulnerabilities.
• Internal System Access: Assess risks before granting open AI access to internal systems.
• Reliability: Exercise discretion in trusting AI-generated content. AI is not infallible and is only as good as the information it has exposure to and the human who give it feedback.
• Misinterpretations: Be aware of AI "hallucinations" or confusion (yes, hallucinations are a real thing in AI).
• Capacity Limitations: Understand and work within AI's capacity constraints (expressed in techspeak as “token limitations”).
• Costs: Evaluate the implementation and operational costs compared to the potential return on that investment.
• Adoption: Facilitate internal adoption and acceptance of AI systems among employees. AI will return the most value when it’s paired with humans, not as a replacement for them.  
• Ethics: AI is a tool that can be used for good or bad purposes. Be sure to foster an ethical approach to AI in your organization.  

By discerning AI capabilities, envisioning potential use cases, and acknowledging associated risks, business owners can make informed decisions about integrating AI effectively into their operations, optimizing productivity and efficiency while mitigating potential drawbacks. It’s an exciting time for AI!

‍