dreman9999 wrote...
AlexMBrennan wrote...
Nope, I don't trust her. EDI was built by Cerberus, so I "trust" her as long as Shepard's goals and Cerberus' align. EDI did not "prove herself" during the suicide mission - because it's exactly what she'd have done if she was working for TIM first*
In particular, EDI being in control of the Normandy is the main reason I destroyed the Collector base - because EDI could easily have killed Shepard once back on the ship, this was the only way to ensure it wouldn't fall into Cerberus' hands.
* You can trust someone when you know that they won't betray you when it would be to their advantage
Oh my god. You do understand she was unshaked then and didn't have to help?
If an AI has all the freedom to rewrite her programing, what stopping her from rewriting it's programing to not help you?
EDI being unshackled is completely irrelevant; Bayesian statistics, however, is not.
Let's, for the moment, assume that EDI had a secret mission: Capture the Collector base for Cerberus at any cost, killing Shepard if necessary.
Would you expect any changes in her behaviour? No, you wouldn't - you'd expect her to help you track the Cerberus cruiser, navigate the Omega-4 relay and all that stuff because her mission and Shepard's (stop the Collectors) align.
As a result, your posterior odds for trusting EDI are unchanged.
Once you blow up the base, EDI has nothing to gain from breaking cover, so again, her behaviour will be exactly the same. Therefore, at the start of ME3 I still have no less reason to believe that EDI has a secret Cerberus mission then when Shepard first met her.
Add the cerberus base videos pretty much proves she is not with cerberus.
That's a video you get to see at the very end of ME3. You cannot use this information to justify Shepard trusting EDI before then.