Building a Practical AI Strategy: Lessons from BRI
Following is an excerpt from an article Adam Noble - vice president, Noble Plastics - drafted for Plastics Business magazine. Visit this link to read the full article.
Introduction
Blue Ridge Industries (BRI) is a custom contract manufacturer specializing in injection molding, high-volume automation and engineering design – creating quality solutions since 1985. The company was founded by technical owners who excel at creating quality solutions and this technical, data-driven culture only has grown over the last 40-plus years. BRI has been exploring the use of artificial intelligence (AI) tools since 2022.
BRI’s AI “Policy”
BRI currently does not have a formal, lengthy AI policy. Technology and use cases are moving so fast that small- and medium-sized companies may be challenged to create and sustain a detailed AI use policy. Instead, BRI team members rely on three guiding principles:
No Non-Public Data Use: Never enter any non-public BRI, customer, supplier or employee data into any free AI tool without a formal, signed data security contract. Does this principle set BRI back? Perhaps, but BRI’s commitment to all stakeholders’ data privacy comes first. In practice, this principle has not harmed the business, as team members have found many other uses for openly available AI tools that provide value and require no proprietary or personal inputs.
Check Accuracy: Review all AI outputs for accuracy before use. If team members use AI-generated content, they must have a human subject-matter expert curate it before use. Blindly copying and pasting AI content can cause serious problems.
Avoid Overuse: Reduce AI dependency and loss of authenticity. There is so much generic, near-useless AI content online right now that it can be extremely off-putting. Being real in an increasingly computer-generated world has immense value.
Strategy
There are so many AI tools available right now that selection can be overwhelming. Choosing the best AI tools is as important as, if not more important than, the applications one chooses to use them with. The landscape is changing rapidly, but right now, BRI’s experience shows that Perplexity is an AI tool that provides citations for sources and answers that one can research more in-depth later. Claude seems especially well-tuned to generating Python, VBA and other code. And of course, OpenAI’s ubiquitous ChatGPT is a well-rounded AI tool that excels in many areas, and team members find themselves using it often.
But there must be more to strategy than just writing good “prompts” for generative AI, right? (Although Google that topic because it is important!) The next layer, and what could separate early users from power users, is how to combine the best AI tools together. One may recall, as a kid, playing Street Fighter arcade games at the local pizza hangout. The simple big-button press provided a nice punch, but for the bosses, combos were needed! Once those combos were known, there was no going back. Is there a way to use one AI tool’s capabilities to feed another that can create something the first cannot? Could AI be used to write the automation for a program? These ideas are explored further in how BRI solved problems using unique AI tool combinations.
For the full story including execution and impact visit Plastics Business magazine.
