Over the past 18 months, there has been a dramatic investment in and scrutiny of the latest AI tools, particularly Generative AI. The technology is nothing short of breathtaking, and history has shown that companies that master new technologies most often gain a competitive advantage and move the needle on market success.
However, increasingly, feedback suggests that "early experiments have been a distraction" and that businesses are not impressed by the results. Why is this happening?
The Allure and Pitfalls of Early Success
The initial ease of using AI can be misleading. Almost anyone with a keyboard can take a collection of data, connect to an LLM like OpenAI, and see some exciting results. This early success can be deceptive. The distance between initial success and a game-changing implementation is a marathon—one many aren't prepared to complete, especially when the first mile was so easy.
What Does This Marathon Consist Of?
Building a game-changing application is not a sprint. There's a lot of groundwork to cover, and every step is crucial to reaching the finish line. The marathon consists of:
- Developing Proper Use Cases: Just because you can build something doesn't mean you should. It's essential to determine the real objectives and problems you're solving for. This step involves identifying the use cases of highest business value.
- Systems Integration: Your data likely resides in multiple systems, tools, or tech stacks. Integrating these systems is a key part of the marathon.
- User Onboarding & Adoption: This isn't just about getting users to start using the system; it's full-on change management. Ensuring that users are comfortable and proficient with the new system is essential.
- Establishing AI Policy and Legal Review: Companies need to create robust AI policies and conduct thorough legal reviews. Data security, both internal and external, also falls under this category.
And this is the easy part ...
The Hard Part: Content Readiness
Now, we move on to the challenging part: content readiness. Is your content prepared for your AI? This question isn't just about the data itself but also about whether the content is suitable for your team, group of users, role, or business unit. Your AI will only be as effective as the quality of the available content. For many organizations, data is often messy and not as well-organized as intended. I recently wrote a blog on this topic: You are what you Eat, feeding your AI.
Combining Talents for Success
Even though these are technology initiatives, technology alone won't guarantee success. I recently spoke with a friend excited about what IT was going to build for their entire organization. My questions were:
- When will IT build solutions for each department, such as sales, operations, the supply chain, and HR, all of which have unique needs and data sources?
- Who will be responsible for training the AI, educating users, and supporting the software in a rapidly evolving environment?
- Who is driving the strategy? Is it a business initiative supported by IT, or is IT leading the charge with everyone else waiting to see the results?
Here's the thing: everyone is new to this. Successful AI initiatives require the expertise and talents of IT and tech partners combined with the domain knowledge and strategic vision of business teams. This marathon requires collaboration, determination, and a clear vision to cross the finish line successfully. When these elements come together, businesses can achieve game-changing results from this latest wave of technology.
