Choosing the Right AI Play
Key Points
- The speaker likens AI implementation to sports, positioning himself as the “captain” who chooses the right AI “play” based on the specific business situation.
- Although Generative AI (GenAI) dominates current buzz, it isn’t the best solution for every problem; using it inappropriately can lead to missed opportunities, higher costs, and brand damage.
- Successful AI projects start by matching the business problem to the most suitable technology—whether that’s GenAI, existing machine‑learning models, or traditional tools—to maximize ROI and minimize total cost of ownership.
- When evaluating any AI solution, four key factors must be considered: the actual need for GenAI versus existing models, the organization’s capabilities to deploy it, how it will integrate with current IT systems, and (implicitly) the governance/maintenance aspects of the chosen approach.
- Leveraging prior AI investments by first exploring use cases that fit those technologies helps optimize overall spend and ensures the right tool is used for the right job.
Sections
Full Transcript
# Choosing the Right AI Play **Source:** [https://www.youtube.com/watch?v=C1ecR-sIE1A](https://www.youtube.com/watch?v=C1ecR-sIE1A) **Duration:** 00:03:40 ## Summary - The speaker likens AI implementation to sports, positioning himself as the “captain” who chooses the right AI “play” based on the specific business situation. - Although Generative AI (GenAI) dominates current buzz, it isn’t the best solution for every problem; using it inappropriately can lead to missed opportunities, higher costs, and brand damage. - Successful AI projects start by matching the business problem to the most suitable technology—whether that’s GenAI, existing machine‑learning models, or traditional tools—to maximize ROI and minimize total cost of ownership. - When evaluating any AI solution, four key factors must be considered: the actual need for GenAI versus existing models, the organization’s capabilities to deploy it, how it will integrate with current IT systems, and (implicitly) the governance/maintenance aspects of the chosen approach. - Leveraging prior AI investments by first exploring use cases that fit those technologies helps optimize overall spend and ensures the right tool is used for the right job. ## Sections - [00:00:00](https://www.youtube.com/watch?v=C1ecR-sIE1A&t=0s) **Untitled Section** - - [00:03:04](https://www.youtube.com/watch?v=C1ecR-sIE1A&t=184s) **Identifying Repeatable AI Use Cases** - The speaker urges spotting repeatable patterns in data architecture, treating use‑case selection like a strategic play that must adapt to real‑time conditions, and points listeners to AI Academy resources for deeper learning. ## Full Transcript
[Music]
In football you have a quarterback, or in proper football you have your playmaker, your #10.
Even a curling side has a skipper.
In any team sport there is somebody on the field organizing the approach and calling the shots,
...based on the specific situation in that moment – Xs and Os, 1s and 0s.
What’s the right play?
Welcome to AI Academy.
My name is Nicholas Renotte. I’m the Chief AI Engineer at IBM Client Engineering.
And if you think of implementing AI as a sport, that’s my job. I’m the captain.
I look at the circumstances of a specific business use case, assess the reality and call the play for the team.
What’s the right AI approach to get my clients the win?
And winning in business means getting better outcomes and the best return on investment.
GenAI is the hottest topic in business right now.
But spoiler alert, it’s not always the right play for every use case. And I’m going to tell you why.
Right now the world is hyper-focused on Generative AI.
But if you’re implanting AI for your business, then you really need to think about your use case and whether it’s right for GenAI,
...or whether it’s better suited to another AI technique or tool.
Otherwise, you could be calling the wrong play or actively working against your team getting that big win.
In business speak, that means missed opportunity, money lost and brand damage.
That’s why it’s so important with AI that you’re implementing it correctly.
And that starts with matching the business problem to the right tool, technique or solution, AI included.
For example, lots of businesses want to generate a financial forecast.
But that’s not typically going to require a GenAI solution, especially when there’s models that can do that for a fraction of the cost.
And other businesses want to use GenAI for optimization, but there’s other well-established patterns which don’t require as much work.
That’s a really important thing to understand.
You need to avoid the temptation to use GenAI where existing tools already work well.
Now you do need to be thinking about and investing in GenAI and AI in general.
But you need to think about tech to problem and problem to tech.
And what I mean by this is that when you have a business problem, you find the right tech to solve it.
And if you’ve already invested in GenAI, AI or machine learning technology,
...you explore the use cases that can be addressed, using your existing investments.
This is the key to optimizing total cost of ownership and ROI for your AI.
When you’re figuring all this out and working through which tech to match to what problem, you need to keep four considerations in mind.
First, needs.
Do you really need GenAI for this project or can your existing ML or AI technologies do the trick and offer a lower total cost of ownership?
Second, capabilities.
Do you have the right capabilities in place for this?
Third, integration.
If you adopted the technology, how would it work alongside your existing IT stack?
And fourth, skills.
Do you have the right mix of skills to take advantage of GenAI here?
Ultimately picking the right use case for GenAI, AI and machine learning tools requires paying attention to a lot of moving parts.
You need to make sure the best technology is solving the right problem.
Be on the lookout for repeatable patterns.
Consider your existing data architecture and look at what’s happening elsewhere in your business.
And you need to do all of these things at the same time.
That’s why finding a use case is like making the right call for a play on the field.
You go out there with all the information you can gather,
...but you need to be prepared to assess conditions on the ground and make changes as you face reality.
That’s how you win.
Check out the AI Academy archives for deep dives into other aspects of AI for Business and watch this space for future episodes.
[Music]