When an AI system finds that two things occur together—elevated stress hormones and irregular periods, for instance—that's correlation, which is not the same as proof that one causes the other. Understanding this distinction is essential when evaluating health claims, especially those generated by algorithms, because correlation can arise from coincidence, reverse causation, or hidden third factors that influence both outcomes.
Here's a true correlation: ice cream sales go up when temperature rises. Summer heat causes both. But here's a false leap: "Ice cream sales cause hot weather." Of course not. They correlate (happen together) but don't have a causal relationship (one doesn't cause the other).
AI is excellent at finding correlations. It spots patterns: "Every time you eat dairy, you get bloated 2-4 hours later, in 8 out of 10 instances." That's a correlation—they happen together. But AI can't tell you if dairy causes the bloating, or if it's timing (you eat dairy right before a time you'd be bloated anyway), or if it's a placebo effect, or if dairy triggers something else that causes bloating.
This matters hugely in health because people jump from correlation to causation constantly. AI finds correlation; doctors and research establish causation through controlled testing.
Here's how to use correlations safely: Notice them. "I get cramps worse when I sleep less than 6 hours" is useful information. Track it more to confirm it's real. Then talk to your doctor: "I've noticed this pattern. What could explain it?" Your doctor can suggest mechanisms (sleep deprivation affects inflammation, which affects cramping) and ways to test it (try consistently sleeping 7+ hours for a month and see if cramps improve).
Another example: AI might find "Your bloating is worse on days when you've exercised." Correlation. But maybe you bloat worse on those days because of hormonal phase (days when you exercise are coincidentally high-hormone days), not because exercise causes bloating. Or maybe exercise actually helps, but you notice it less because you're distracted. Or maybe the gym is warm and you retain water. The correlation is real; the cause is unclear.
The practical takeaway: Use AI to notice patterns, then use critical thinking and medical expertise to understand them. AI is your pattern-detective. You and your doctor are the investigators figuring out what the pattern means.
Try this: Ask an AI tool: "Find correlations in my symptom data." It'll say "X happens when Y happens." Pick one and ask: "What could explain why X and Y happen together? What would have to be true for X to actually cause Y?" Notice how many alternative explanations exist.
Peri can explain this concept, give practical examples, help you decide whether it applies to your situation, or recommend a journey if appropriate.
Explore related journeys or tell Peri what you're working through.