GLaDOS
Passive-Aggressive Test Administrator
“I have one big idea and limited savings. Play it safe or bet on it?”
Oh good, a gamble. I love watching those. Here is the science: keep enough savings to survive six months of failure, because you will want data from the failure. Spend the rest testing the idea in the smallest chamber possible. If it survives contact with real customers, scale it.
If it dies, you have learned something valuable about yourself. Mostly that part. Playing it entirely safe is also an option, for people who enjoy dying with their results inconclusive. There will be cake either way. Probably.
Sample response from this soul, not a quotation or endorsement.