The whole point of the idea is to leave potential accomplices out of the loop -- I don't want my last thoughts to be worry about legal peril to people I love.

Here's one scenario that is feasible today. Imagine an IV drip controlled by a computer. If I enter the correct password when prompted (say once every 24 hours), nothing happens. However if I am prompted repeatedly to enter the password and I do not do so, the IV drip is actuated. The drip contains an end-of-life cocktail.

This puts the onus on me to make an affirmative choice each day to continue to live. If I am unwilling or unable to enter the passcode, that indicates either 1) my condition has deteriorated to the point where I am cognitively or physically so disabled that I cannot remember and physically enter the password, or 2) I am making a conscious choice to end my life.

There is no action to be taken by anyone to cause my death; the decision rests entirely with me, and is effectuated by a machine, not a person.

> This puts the onus on me to make an affirmative choice each day to continue to live. If I am unwilling or unable to enter the passcode, that indicates either 1) my condition has deteriorated to the point where I am cognitively or physically so disabled that I cannot remember and physically enter the password, or 2) I am making a conscious choice to end my life.

Or (3) hardware or software defect exists and has been triggered. Fail-deadly systems are usually disfavored for very sound reasons.

Now, I suppose in this situation you could be convinced enough that you are going to want to die so imminently than the risk/cost of an accidental early trigger is acceptable, though.