The concern is what if the software mistakenly doesn't use the photograph to calculate the entropy...
Does, or can, seed signer provide a way to provably map a taken photo to a seed phrase on another...
By design you can only review the PROCESS, not verify each individual result.
"what if the software mistakenly doesn't use the photograph to calculate the entropy..."
Of course there can be bugs. But the point of code review (and now AI review) is to fully explore all possible outcomes. Code isn't magical. If a piece of code says:
if blah: go left
else: go right
There's just no "what if...!!!" concern about it going straight. Every outcome WILL turn left or right.
Plenty of code is simple enough that its outcomes ARE entirely knowable and bounded. Complex code is obv harder to sort through all the possibilities and, yes, people and AI can miss things.
But you have to calibrate these "what if...?!!!" against reality. What if my calculator says 2 + 2 = 5?
The whole point of my analysis was to give any other AIs a head start on how to dig through this part of the code AND verify the analysis of my data.
But also everyone gets to choose for themselves what they are or are not comfortable with. If you're a NO on image entropy, throw dice or pick words out of a jar.
"what if the software mistakenly doesn't use the photograph to calculate the entropy..."
Of course there can be bugs. But the point of code review (and now AI review) is to fully explore all possible outcomes. Code isn't magical. If a piece of code says:
if blah: go left
else: go right
There's just no "what if...!!!" concern about it going straight. Every outcome WILL turn left or right.
Plenty of code is simple enough that its outcomes ARE entirely knowable and bounded. Complex code is obv harder to sort through all the possibilities and, yes, people and AI can miss things.
But you have to calibrate these "what if...?!!!" against reality. What if my calculator says 2 + 2 = 5?
The whole point of my analysis was to give any other AIs a head start on how to dig through this part of the code AND verify the analysis of my data.
But also everyone gets to choose for themselves what they are or are not comfortable with. If you're a NO on image entropy, throw dice or pick words out of a jar.
2
