When AAA Insurance meets with Joe Sixpack to discuss his health or life or earthquake insurance, they seek to collect enough information to have a reasonable chance of turning a profit on the deal. Else why would they exist as a business?
But there is a necessary asymmetry here. If you could determine with good precision whether someone will be affected by an illness, when, how much, etc, then this wouldn't work, save for superstition on the part of Joe Sixpack. Since the contract is based on a bet on the likelihood of premium/payouts balance, the more you can find out about the future of the insured person's organism future, the closer the premiums will match the payouts, reliably. *IF* you can determine this, of course. So the goal will be for the insurer to get access to as much info as possible to assess how to set the premiums, while preventing the insured from knowing as much, so there is still the uncertainty and the value of "peace of mind" gotten by the insured, and that's worth something too. The converse is also true. And both raise the problem of assessment of the data - how does the insurer get the data (getting DNA from the would be insured ? with the insured's knowledge or not ? From the contract clause that subordinates the insurance to the supplying of the data by the insured ? Will credit bureaus expand to cover this kind of thing ?) If there is total symmetry, insurance loses its point entirely. Could we see a gradual disappearance of some sorts of insurance for events that cease to be probabilistic ? Just musing. All of it already happens, but I'm curious about the limit of it (in the mathemetical sense) when the precision of the prediction tends towards infinity. -- Vincent Penquerc'h