Why is enterprise AI becoming harder to deploy, even as large models grow more powerful?
Over the past two years, companies invested heavily in AI but often faced an awkward result: AI launched, yet business didn't change. Models are capable enough; what's missing is a way to connect AI to real operations. For manufacturers predicting equipment failures, the model is only 20%—the other 80% is data, system integration, maintenance know-how, and process complexity. This gives rise to th
AI Is Live, but the Business Hasn't Changed: What Enterprises Lack Is Not a Smarter Brain, but a Way to Connect It
Over the past two years, the money enterprises invested most heavily may have bought an awkward result:
AI is live, but the business has not changed.
Many companies now have large models.
Employees have started using AI to write documents and do analysis.
But the questions the boss truly cares about:
Has sales grown?
Has production become more efficient?
Have costs come down?
Many still have no answers.
Is it because AI isn't strong enough?
Quite the opposite.
Today's large models can already accomplish many tasks that were previously unimaginable.
The real problem is:
What enterprises lack is not a smarter brain, but a way to connect that brain into the company.
For example, a manufacturing enterprise wants to use AI to predict equipment failures.
It sounds simple.
Just let AI analyse the data, right?
But when it comes to actual deployment, you discover:
Where is the equipment data?
How do you connect to the MES system?
Has maintenance experience been accumulated?
How do process changes affect judgment?
In the end, you will find:
The model is only 20%.The remaining 80% is all about enterprise on-site problems.
01FDE: A New Role Emerging in the AI Industry
This is also why a new role is emerging in the AI industry:
FDE, Forward Deployed Engineer.
They do not sit in an office developing a generic product.
Instead, they go directly into the enterprise site:
understanding business processes,
connecting enterprise systems,
transforming ways of working,
and ultimately making AI truly deliver value.
In the past software era:
the competition was about who developed more software features.
But in the AI era:
the competition is about who better understands the enterprise's real problems.
Because in the future, model capabilities will become increasingly widespread.
What is truly scarce is:
the people who can turn AI capabilities into business results.
This is also why AI companies such as OpenAI and Palantir are placing increasing emphasis on the FDE model.
In the next issue, the discussion continues:
Why FDE may become one of the most important new professions in the AI era.