In this installment of “Artificial Intelligence for Fun and Profit” I’ll discuss small business use. The biggest advantage to adopting AI-powered systems is to increase productivity and access capabilities you don’t necessarily have in-house. Additionally, AI is great for automating mundane tasks.
The first thing you have to get used to is using a prompt, which is how you ask AI to do something for you. It’s essentially how you “program” AI. You use plain English, and the more specific you are, the more specific the result. However, it may take several tries until you get it down.
Considering I run an industry blog, the biggest question I get is about using AI to write press releases. I’ve covered the topic of best practices for press releases in the past, and I suggest you review it on the Tactical Retailer website (tacretailer.com; search “Graves”). You can absolutely use AI to prepare effective press releases, but I can’t reiterate it enough: Check the work. It may require some tweaking to tell your story accurately.
One of the main tells I see regarding AI-created PRs is that the title contains punctuation, making it a sentence rather than a bullet. I see loads of PRs every day, and between that and voice, or how the piece is written, I can detect AI use.
This content creation can be augmented with the creation of custom images and even entire videos, including voiceovers using generative AI platforms like Canva’s Magic Studio, Adobe Firefly, or Jasper. The results can be used as content for social media campaigns that can be further managed by AI-powered tools in Brandwatch, Talkwalker, or Hootsuite Insights to target potential customers.
Many people aren’t comfortable writing, particularly about themselves and their business. AI can handle the most mundane writing tasks. Some examples include “About Us” statements, employee bios, and even business plans.
Another use of AI is market research. Not only can you learn about your market competitors, but you can also research new products. Large Language Models are useful for such research. If you value data privacy, cost-efficiency, and alignment with traditional values, there are several LLMs you might consider.
Although Llama 3.1 is from Meta AI, it can be self-hosted for those who are avoiding cloud-based services, which helps protect your data. Due to its open source, it can also be customized via fine-tuning with company-specific data. Its small size runs very quickly but requires setup. Free to download from GitHub for research and commercial use, its use is limited to businesses with fewer than 700 million users. Llama is well suited for customer support chatbots, internal document summarization, and localized content creation.
If your concern is freedom from programmer bias, you might consider Grok-3 from xAI, which emphasizes truth-seeking and minimal ideological bias. Web-based, this is great for research and analysis.
If you’ve ever interacted with a chatbot online, the initial interaction was via an AI-driven agent, which responds to customer prompts to provide the answer to their questions. Naturally, you’ll need to train your chatbot about your company, products and how you do business, but it learns quickly and doesn’t forget. A couple of chatbots you can use to automate interactions via social media include ManyChat or MobileMonkey for Instagram or Facebook Messenger.
One investment that will pay off is in customer surveys. The more you learn about your customers and their spending habits, the more you can customize their shopping experience, both online and in store.
Ever had products suggested while checking out online? The business has compared what’s in your cart with others who have purchased the same items and suggested you might want them as well. It takes those checkout-line impulse buys to a new level. Likewise, you can prompt a customer to buy complementary items for their current purchase, either by bundling or via suggestion by a salesperson in-store.
You’d be surprised at how many software applications and apps you already use rely upon AI. For instance, Intuit Quickbooks automates bookkeeping, categorizes transactions, and generates financial reports. It can also predict cash flow and simplify tax preparation. Additionally, Microsoft Copilot is an AI-powered assistant integrated into Microsoft 365 apps like Word, Excel, PowerPoint, Outlook, Teams and more. It leverages large language models (LLMs), including OpenAI’s GPT-4, combined with Microsoft Graph to provide context-aware assistance.
Additionally, you can use AI apps to make records of meetings, offering either complete transcripts or summaries and to-do lists by participant based upon what was discussed.
While I may have mentioned a few commercial product names, I am by no means endorsing them, but rather passing on what I, or others have used. The beauty of AI is that you can ask it to help you select the right product for your application, adjusting your prompts as you narrow in on the right choice.
AI is becoming more common as it integrates into more and more aspects of our lives. This topic is also complicated, and I’ll dedicate several articles to covering it. In the future look, for additional installments discussing government/military applications of AI, the importance of data and, finally, future business opportunities resulting from AI.
Like in the first installment, I asked Grok to create an image for this article. My prompt was “create an image of gun store use of AI.” It gave me two options, and I chose this one. The other featured a firearm with barrels pointed in both directions, which takes me back to the most important point in using AI: Always check the work.
In closing, I’d like to disclose that I used AI to draft this article for research. I do so quite often in my day-to-day writing, particularly when researching technical details and seeking out the latest information. Everything remains written in my voice and analyzed and presented by me.
