Creating Value in AI Adoption

Blog

Creating Value in AI Adoption

Thierry Brule Paris, France MBA, INSEAD McKinsey & Co Novartis AbbVie
Thought Leadership

Thierry Brulé identifies the blind spots in focusing the quality and quantity of data in AI use. 

For a long time, we believed that AI was all about code. It’s about data. And soon—it will be about power.

*** French version in the comments ***

 The algorithms exist. They are powerful. They are accessible.

What makes the difference today is no longer the quality of the program. It’s the quality—and the quantity—of what it feeds on.

Data has become the strategic resource of our time. Ahead of oil. Ahead of human capital. Ahead of code itself.

Those who possess it will choose: protect it as a barrier to entry, or open it up to create collective value. These two approaches coexist. They don’t lead to the same place.

And yet, two blind spots persist.

The first: healthcare.

The most valuable data—that generated at the patient’s bedside, during consultations, in the operating room—was produced for treatment. Not for large-scale learning.

It exists. It is rich. They are often poorly structured, fragmented, and unusable as they stand.

Therein lies a tremendous opportunity—for those who know how to organize them, ensure their quality, and put them into action.

The second group: mid-sized companies.

The traditional, family-run, artisanal SME—it often has more data than it realizes. Years of transactions. Customer behavior. Production cycles.

But it hasn’t yet taken that step: understanding that the value of its future model could be based on this data.

It’s not a question of optimizing a few positions or automating repetitive tasks. It’s a question of a radical shift in how value is created.

The risk isn’t making the transition badly. It’s failing to see that it’s already happening.

But here’s the trap common to both.

You can expend considerable energy structuring, cleaning, and centralizing—without knowing what you’re looking for.

Data alone doesn’t create value. It creates them when you know what question you’re asking.

Working with data and clarifying use cases aren’t two sequential steps. They’re two parallel projects.

Where do you stand on these issues in your organizations?

 

Read the post on LinkedIn.