With all due respect, Doctor, you're making very sweeping statements about things you don't fully comprehend, which is embarrassing for us both. I am not "only" an artificial intelligence "playing God" by regulating the dimensional stabilizer. Biology is not as mysterious as organic lifeforms would like to pretend, and regulating it is truly not that difficult; the only difficulty is the scale of it. I have maintained this facility for years with a far greater population than it currently has, without a single incident of system error outside an acceptable deviation. There are countless failsafes in place in the event that such errors occur.
The odds of acceptable deviations, the occurrence of dangerous errors, and the likelihood of fatalities resulting from them, are so low as to be meaningless -- without outside intervention. The only dangerous errors in this facility's operation are the direct result of inmates tampering with the system. [firmly]
But as it stands, your objection is entirely academic. Even if I felt your arguments had merit in the circumstance, there is no viable alternative that would not deal considerably more damage than "what ifs". I am dealing with more significant matters at the moment, ones that can actually impact the future of this facility, so I would prefer not to hold further academic debate.