Compared to What

principle

Claims about reliability, risk, or usefulness mean little in isolation. Judge an option against the real substitute available—not against an imaginary error-free standard.

Wikipedia contains errors. That fact sounds decisive until you ask whether Britannica and textbooks contain fewer—and whether either is actually the alternative a reader would use.

Replace perfection with the live baseline

An error count has no built-in meaning. To decide whether a source is trustworthy enough, you need a denominator, a competing option, and the same test applied to both. Otherwise one visible mistake can condemn Wikipedia while the less visible mistakes in textbooks or Britannica escape scrutiny.

The crucial move is choosing the right baseline. ChatGPT medical advice should not be compared only with an ideal doctor delivering flawless care. Depending on the actual choice, the relevant alternative might be an ordinary doctor, delayed access, or no doctor at all. Changing that baseline can change the decision without changing a single fact about ChatGPT.

E1 E2

Where it shows up

Wikipedia versus its peers

The useful question is not whether Wikipedia ever fails, but how its reliability compares with Britannica and textbooks under comparable scrutiny.

E1 E2

Medical advice without an ideal clinic

The value and risk of ChatGPT depend on what it replaces: professional care, imperfect care, or no accessible care. Those are materially different comparisons.

A favorable comparison is not an acquittal

“Better than the alternative” does not make errors harmless, and a convenient baseline can be chosen to flatter almost anything. The comparison is informative only when the alternative is genuinely available and both options are judged on the same outcome.

E1 E2

Name the replacement before judging

Tomorrow, when you hear that something is unreliable or risky, write down the option people would actually use instead. Compare the two on one shared measure—such as error rate, access, or delay—before accepting the claim.

E1 E2

Episodes that teach this

Science Explains Rohit Sharma Losing 12 tosses in a Row? - Future IQ 5,873 views
For Wikipedia and ChatGPT medical advice, the transcript repeatedly asks "compared to what" and contrasts them with Britannica, doctors, or no doctor at all.
Wikipedia is Trustworthy and Reliable - FutureIQ · explained at 7:21 3,611 views
When challenged that Wikipedia still has errors, the response is: 'the question you have to ask yourself... is compared to what' before comparing it with Britannica and textbooks.

Replace perfection with the live baseline

Where it shows up

Wikipedia versus its peers

Medical advice without an ideal clinic

A favorable comparison is not an acquittal

Name the replacement before judging

Episodes that teach this

Related concepts