Recovered data from the DSMP simulation project:
This discussion seems to take place immediately after recovered data #34.

BEGINNING OF RECOVERED DATA:
[DATE REDACTED]

Dr A: Do we take out Subject K in prevention?

Dr B : We can try to erase his latent memory data and then put him back in the simulation

Dr A: We'll need to unplug him to do that

Dr B: Then we terminate him, he's the most dangerous one.

Dr A : So we terminate Subjects S and K and we put Subject D in rehabilitation ?

Dr B : Yes we can do that for now.
It's the safest way, we can't send them back and they could escape if they stay in the facility.

Dr A: Okay I'll send a team unplug Subject D
We need to put him in rehabilitation as quickly as possible so he doesn't notice something is wrong.

[CORRUPTED DATA]

Dr A: The problem is everyone else is in the same simulation right now, they're gonna notice those three are missing, we'll need to run a new simulation.

Dr B: Or maybe put other subject in the simulation
Some who are identical to S and K.

Dr A: Yes ! Data manipulation ! We can send AI substitutes and make them believe it's them!

Dr B: But we have to be very careful, unless they realize the substitutes are fake.

Dr A: With subjects K, S and D out of the picture, the only thing that had any contact with the fourth wall was a basic figurant AI , we can recalibrate it.
But yeah, data manipulation on an ongoing complex simulation with multiple subjects is very delicate

Dr B: It should be ok, if everything goes according to plan.

END OF RECOVERED DATA