Companies approve AI budgets before asking what they’ve built. Demos look impressive, vendors promise transformation. Then the silence – as executives can’t explain the ROI. We’ve overcomplicated it.
We’ve overcomplicated it. There are two Lego bricks of AI that deliver business value: Assistive and Generative. Everything else is just a Lego construction.
Assistive AI: Find What Already Exists
Assistive AI brings to the surface information that’s buried in data. “Tag the faces in this footage”, “Find the car in this frame”, “Extract metadata from this asset library”.
The output is text, tags, or structured data that lets you navigate what you already own.
Think of it as hiring the world’s best librarian. Your data doesn’t change, but suddenly you can find the right stuff. Media companies who’ve spent decades accumulating footage – Assistive AI is the first technology that makes archives financially viable to search.
Generative AI: Creating Something Out of Nothing
Generative AI makes new things. “Make an image from this description.” “Create a video from two images and a prompt” The end result is data that didn’t exist before the prompt. You’re using AI models to generate assets, not locating existing ones.
Where Assistive AI lets you leverage existing datasets. Generative AI rapidly explodes your data generation – under the illusion of becoming “more efficient”.
The Third Category Is Just Marketing
Agentic AI sounds revolutionary until you look at what it is. Chains of Assistive and Generative processes, combined with business logic. Anyone from the Media & Entertainment will know these as what they really are; Workflows.
The terminology is different because we’re in the era of hype. All we’re doing is replacing steps in our businesses processes with AI-powered black boxes.
What This Means for Your Organization
The strategic question isn’t “How do we use AI?” – it needs to be “What problem are we solving”
If the answer is “Findability”, then Assistive AI will be in play. If the answer is “Production” then it’ll be Generative AI.
Most media organisations may have a production challenge that’s actually a findability problem at its core. Commissioning new content while petabytes of existing footage sits unmanaged is a common theme across the industry.
The Real Implementation Decision
AI is already reshaping your organisation through shadow AI workflows that teams use every day. The choice isn’t whether to adopt, it’s whether your budget, strategy and competitive position gets dictated or derailed by the next AI productivity tool of the month.
The companies that win won’t have the best AI models. They’ll see their data problem clearly enough to choose the right type of AI, and see it in time to shape it.