As an Oncologist observing this, This may ultimately prove to be one of the most important applications of AI in oncology.
For decades, we classified cancers by where they arose:
Breast.
Lung.
Colon.
Pancreas.
Then we learned to classify them by mutations.
Now we are entering a third era:
Classifying cancer by cellular state.
Two patients may carry the same diagnosis.
The same mutation.
The same stage.
Receive the same drug.
Yet have completely different outcomes.
Why?
Because cancer is not a static photograph.
It is a living, evolving ecosystem.
Every tumor contains millions of cells making different decisions in real time.
Understanding those cellular states may help explain why some patients experience extraordinary responses while others derive little benefit.
This is where AI becomes transformative.
Not because it replaces physicians.
Not because it replaces scientists.
But because it can recognize patterns across billions of biological interactions that no human mind could ever process alone.
The future of oncology may not be asking:
“What cancer does this patient have?”
But rather:
“What is this cancer doing right now?”
That shift is profound.
If we can understand cell state, predict resistance before it occurs, and match therapies to the evolving biology of each patient’s tumor, precision oncology becomes truly precise.
The greatest challenge in cancer medicine has never been killing cancer cells.
It has been understanding them.
AI is giving us a new language to read the hidden conversations occurring inside tumors.
And once we can read those conversations, we may finally learn how to change their ending.
The future of cancer care will not be built by biology alone or computation alone.
It will emerge from the convergence of both.
And that future is arriving faster than most people realize.