Why Context Is A Differentiator When Companies Implement AI

​Marc Potter is the CEO of Actian.

AI has become increasingly available to companies and is transforming businesses. When companies all have access to the same models and AI technology, the real differentiator is how your organization can uniquely apply AI. It’s not about the data. It’s all about the human layer: the contextual knowledge that your employees have about the business and your customers that the data alone cannot capture.

In my conversations with customers, there is a debate over the context layer and where it should “live” in an organization. It can exist in systems, data, agents or within your employees.

People understand the who, what, when, where and why behind the data. They know the history and relationships between people, systems and data that are not interconnected by digital systems. Their undocumented knowledge enables them to identify anomalies. With this insight, they can help train the AI solution so it can understand the context around the data, provide proper guardrails and governance and determine the appropriate next steps rather than guessing.

Giving AI Guidance

The human layer of context is frequently overlooked when scaling AI across an enterprise, and its influence goes well beyond what someone types into a prompt. AI requires guidance such as understanding its purpose and the company’s objectives, the company’s philosophy on customer experience or the company’s rules about data usage.

AI should be treated like a non-human employee (NHE) that works alongside people and technology to accomplish a common goal or outcome. Training AI should be approached like you would onboard a new employee to the company, where a human sets the direction and purpose. Not a “human in the loop,” but “human as the lead.”

The Risks Of Using AI To Justify Layoffs

That intellectual property is at risk when leaders use AI to justify layoffs. The people leaders remove may possess precisely the context required to make their AI investments successful.