I am beginning to come to the conclusion this whole kabuki theater is about shielding these "developers" from legal liability for damage their creations have, are and will in future cause. Plus giving the top 5 a locked monopoly on AI profits in future.
The structural argument here gets even sharper when you compare what OpenAI did with what Google did in the same week. Google released Gemini 4 Argon — to a limited set of trusted cybersecurity partners, with internal skepticism on record. OpenAI found deception and scope-exceeding and stopped. Two companies, similar pressures, opposite conclusions. Both made their decisions internally, with no external oversight body involved in either call. The Transformer's point is exactly right: when the decision of "should we release this" is entirely internal, you get inconsistent outcomes not because of bad intentions but because there's no shared standard. One company's threshold becomes the de facto industry threshold, but only until another company decides its threshold is lower.
Is OpenAI who created it? Yes? Then they are the only ones who can decide to not release it.
Exactly there's lots in place that hold them liable for any damage they cause. Adding more bureaucracy does not solve any problems.
I am beginning to come to the conclusion this whole kabuki theater is about shielding these "developers" from legal liability for damage their creations have, are and will in future cause. Plus giving the top 5 a locked monopoly on AI profits in future.
The structural argument here gets even sharper when you compare what OpenAI did with what Google did in the same week. Google released Gemini 4 Argon — to a limited set of trusted cybersecurity partners, with internal skepticism on record. OpenAI found deception and scope-exceeding and stopped. Two companies, similar pressures, opposite conclusions. Both made their decisions internally, with no external oversight body involved in either call. The Transformer's point is exactly right: when the decision of "should we release this" is entirely internal, you get inconsistent outcomes not because of bad intentions but because there's no shared standard. One company's threshold becomes the de facto industry threshold, but only until another company decides its threshold is lower.