What is causality?
When one event (the cause) leads to another (the effect).
Causality describes why things happen and how factors shape outcomes. This differs from correlation, in which two factors may be statistically associated with each other, without necessarily being directly causally related.
In the scenario of dominos falling, the finger push factor causes the event. The fingernail colour factor is related to the event but is not its cause.
What is causal AI?
An artificial intelligence designed to untangle complex causal relationships and to use them in robust problem-solving.
This differs from traditional AI, which identifies patterns and connections in data. Causal AI digs deeper to figure out what drives outcomes. This means it doesn’t just spot links — it understands why things happen.
Why does healthcare need causal AI?
Healthcare is incredibly complex.
Outcomes are dependent on intertwined factors such as genetics, lifestyle, environment, and treatment plans which interact in complex ways. Understanding why outcomes occur can allow us to make smarter, actionable, more personalised decisions, develop better treatments, and save lives.
For example, causal AI can process large amounts of data for experiments, clinical trials, increasing diagnosis/prognosis speed, and tailoring patient treatments.