CompTIA’s Artificial Intelligence Advisory Council has recently created a detailed document outlining the top questions decision makers and practitioners need to ask before implementing AI in businesses. Here’s why.
CompTIA’s Artificial Intelligence Advisory Council members recently created a detailed document outlining the top questions business leaders and AI practitioners should ask before implementing AI.
Rama Akkiraju is an IBM Fellow and distinguished engineer. She also chairs the CompTIA AI Council. She spoke with CompTIA about the council’s intentions, common misconceptions regarding AI implementations, and what decision-makers should do after reviewing the recommendations. Here’s what Akkiraju had to say after receiving the CompTIA Industry Advisory Council Leadership Award.
Why did you create the Top Considerations document?
The Artificial Intelligence Advisory Council at CompTIA is responsible for generating best practice documents for practitioners as well as business decision makers. They will be embarking on AI-infused projects and looking at how AI can optimize various aspects of their enterprises. What should they be doing? What are the best practices? We wanted to create a set of work products to guide them on how to think, ask the right questions, and what it takes to reach the maturity level they desire.
What is the biggest misconception that businesses have about AI implementation?
Some promises are not realistic. AI is being promoted by some companies and individuals as magic. A button is all it takes to get a prediction right 100% of the times. It doesn’t require any maintenance or new skills. It will always be fresh and do the right thing. It will seamlessly integrate with my existing products and all that. Although I may be exaggerating a bit, it is probably not too much.
We wanted to decode that process and show that AI is not magic. But it can transform many industries dramatically. All it takes to make that happen is hard work, data and new skills. It takes a deep understanding of how they work, and then giving feedback. They are not perfect from day one.
How can MSPs improve their AI solution development and implementation skills
They must do something else. Depending on the type of AI being discussed, AI requires preparation, data preparation or expert writing rules. It will require work from the person who is developing it, the person who is deploying it, as well as the person who is managing it.
It’s similar to traditional software application development in some ways but it’s quite different in others. It is important to understand the problem you are solving and to use the software engineering principles of requirements to document it. In other cases, however, it is very different to traditional software engineering in that you will need new skills and roles such as the data manager to prepare data for AI. It could be the operations engineers, who must understand how AI works. They need to understand that it is not always accurate and that it requires iterations, additional data and feedback to improve it. They must also understand the aspects of AI-based solution development so they can build and manage them.
What should businesses do after reading the top considerations document?
First, look at their business processes and ask yourself: “Which of my business processes can I leverage and benefit from AI Infusion?” Where should I start? What are the most important to me for this exploration or proof-of-concept?” Once they have identified these in their business processes, they can start looking at other options.
You may already know of vendors or companies that offer solutions to optimize this business process. You might create a request to be presented in order to get information from all the proposals and presentations. You can then decide whether or not you want to use your internal data.