Artificial intelligence (AI) is fraught with risk. But your small or midsize business (SMB) shouldn’t ignore it. Instead, apply your human brainpower, intelligence, and judgment in this increasingly artificial world.
In our last blog, we explored Scary AI: 13 Frightening AI Facts. In this blog, we offer Mindful Advice for AI in Business. If you’re a business owner or leader, consider these six AI strategies.
1
Understand What AI Is.
First, understand what AI is and how it works. This will help you evaluate AI applications for your business.
Recognize that AI is an umbrella term for several types of computer intelligence. These include:
Machine Learning
AI trains itself!
Ability of computer systems to analyze patterns and make inferences based on algorithms (sets of instructions) and mathematical relationships (statistical models).
Deep Learning
AI makes complex connections and learns from experience!
A subset of machine learning, deep learning goes further than basic machine learning. It uses layers of processing to recognize complex patterns.
Computer Vision
AI sees!
Ability of computers to recognize people and objects in images and videos.
Natural Language Processing (NLP)
AI reads and listens!
Ability to interpret written and spoken words and their relationships.
Generative AI
AI creates!
Uses deep learning to create new content from raw sources. This content can be images, videos, audio, documents, and more.
Natural Language Generation (NLG)
AI writes and speaks!
A subset of generative AI and deep learning, NLG uses software that can speak or write a story from structured data (in tables) or unstructured data such as language, audio files, emails, spreadsheets, presentations, video files, or large volumes of documents.
2
Become and Stay Informed—Especially About the Use of AI in Your Industry.
- Keep up with industry trends. Go online and research the ever-evolving types of AI platforms used in your business discipline. Read objective scientific articles and trade news about AI in your field.
- Learn from experts. Attend conferences, join webinars, and participate in online forums.
- Pay attention to the news. Examine objective, balanced news about AI (not just what AI purveyors promote). Set up Google Alert to track AI news.
- Observe the law. Stay abreast of current and proposed state and federal legislation about using AI.
- Network with colleagues. Reach out to peers in your industry; discover how they use AI and what they’ve learned.
3
Understand which databases were used in the creation of the AI application. They may have inherent bias or be known to hallucinate.
Critique Specific AI Applications.
- Mind their data. Understand what databases and sources were initially used to create the AI application and how it has evolved.
- Mind your data. Know what happens to your inputted data.
- Mind your use. Be aware of what happens to the content you generate.
- Identify potential errors AI may generate. How accurate is the AI application? AI may hallucinate (tell outright lies that sound plausible).
- Consider cost and energy expenditure. Many AI applications have free versions. Know that generative AI uses substantially more electricity.
- Recognize risks AI might create. AI applications may potentially infringe on copyright, breach security, or violate privacy.
- Anticipate the future. Consider what consequences may come next. For example, companies such as Meta (Facebook and Instagram) used facial recognition software, which uses AI, without having express permission to do so. Later, this resulted in statewide class action lawsuits with hefty fines. Meta was ordered to pay $1.4 billion to Texans. Illinoisans received $650 million for Facebook violations and $68.5 million for Instagram violations.
4
Be Proactive.
- Develop company policies. Write an AI policy that incorporate ethical, legal, and privacy considerations. Where do you need to provide customer transparency about your use of AI?
- Evaluate your existing data. Review your relevant data and how it could potentially solve business problems.
- Consider the consequences. Evaluate whether or not you have the right to use available data in an AI application, considering both current and forward-looking security, privacy, ethical, and legal consequences.
- Empower people. Investigate low-risk, high-reward ways to empower your team with AI.
- Understand the complexities. Hire trustworthy data experts who understand bias in the data.
5
Don’t share personal information—yours or anyone else’s—with AI. Don’t use AI for sensitive topics.
Avoid High-Risk AI behaviors.
- Don’t share personal information—yours or anyone else’s—with AI applications.
- Don’t use AI for sensitive topics.
- Don’t generate large volumes of repetitive content, aka spam. Your customers will disconnect.
- Don’t use content verbatim. Copy and paste functions may be tempting but don’t plagiarize AI. First, AI tools tend to write pablum—dry, dull, paragraphs that lack pizzazz. Second, paragraphs are often lengthy and require editing. Third, AI may hallucinate (lie). Furthermore, AI may place you at risk of plagiarism or copyright infringement.
- Don’t make significant life or business decisions based solely on AI.
- Don’t believe everything you read or see.
- Don’t put your head in the sand. Although the recent proliferation of AI tools is exponential, AI has been evolving since the 1950s. It’s here to stay—whether or not you and your business are prepared.
6
Play with AI. Experiment with free versions of AI applications and tools. Left: Original photo of Shorty, the author’s first foster rescue from early 2020. Center: Canva was asked to generate a Japanese anime style. (Look closely, and you’ll notice that poor Shorty also generated either a fifth leg or a second tail. Also, his leash looks suspiciously like an ethernet cable on one end.) Right: Canva was asked to generate the image of Shorty on a raft in the ocean.
Be Mindful.
- Play with AI. Experiment with free versions of AI applications. Make a cartoon version of your dog in Canva. Ask an AI chatbot, such as ChatGPT or Gemini, for birthday party ideas.
- Be specific with your prompts. Be clear and concise, and use natural language. Context also helps. The better your prompt, the more likely you will get usable results. Using the birthday party example, you might ask for surprise birthday party ideas for a 40-year-old man.
- Leverage AI’s strengths. Ask an AI chatbot to summarize data, make lists, or quickly generate ideas.
- Use AI for low-risk, high-efficiency activities. For example, you might use AI to find the best title for your workshop, list all salient points in your email, or garner feedback on your (public-facing) content.
- Implement AI for efficiency or improved productivity on topics you know well. If you’re an expert, you can easily verify facts.
- Check AI’s work. Be sure it’s not hallucinating (making stuff up). If you use AI to generate content for which you’re not an expert, authenticate AI-generated “facts” with peer-reviewed literature or subject matter experts (SMEs). If it generates a citation, be sure to click on the citation and check the source. Likewise, test computer code AI generates. If you use an AI translation tool, ask a native speaker to review and edit the accuracy of the translation.
- Pay attention. Often. It’s easy for midsize and small business owners to get caught up in day-to-day activities. Be sure to step back (e.g., at least quarterly) to see how AI has evolved. Meteoric growth is an understatement. Keep up. This period of rapid growth is likely to continue for another decade.
- Pause and reflect. How could the latest AI advancements help you, your team, or your SMB?
- Don’t mindlessly “believe” AI tools. For example, Grammarly suggests changing “more than x” to “over x” or “such as” to “like.” But “more than” and “such as” are grammatically correct—and its suggestions are not. At times, it suggests phrases that substantially change the sentence’s meaning. However, it also pinpoints keystroke errors and other typos, avoiding the need for a second proofreader.
- Trust yourself. We haven’t achieved singularity yet. Use your judgment. Be mindful. Stay mindful.