I never had an issue with AI up-scaling as it is way more effective and efficient than the legacy methods. I actually am planning to re-construct my AI research server at some point in time to help science progress much faster than it could by conventional means. I use BOINC (Berkeley Open Infrastructure for Network Computing) So yea, there are good uses, neutral uses, and bad uses for AI. I always looked at it subjectively. By the way, thanks for bringing back the legacy Omega drill as I never liked how the new one was designed. Really liked its' looks, but never liked how the mining to its' own inventory was incarnated.