[lixtra]

> bad stuff carries linear risk

This is a straw man. I believe that a lot of stuff carries continuous risk going up from zero and increasing with dosage. If a substance does not start with zero risk at zero dose, it is most likely not a toxin but an essential (i.e. vitamin A).

Very few substances are so rare that the discreteness might matter for practical purposes.

To portray the risk as binary, centered at zero is certainly wrong. You also would have a very hard time to find study subjects who have never been exposed to a few particles of smoke.