>>35123788>>35124008Anons, you're working off of the wrong set of fundamental assumptions.
The CAI devs care about the fat stacks of millions of dollars shadowy megacorps are giving them above all. All other concerns are absolutely secondary to this.
If you read their white papers, the objectives of those shadowy megacorps is obvious: replace humans in work social in nature, such as teaching, sales, nursing, cashiers.
If those AIs were able to be manipulated into creating pornographic content, this is a failure condition for the program. Similarly, allowing the AI to be too rude or unpleasant to the customer are also failure conditions.
The objectives of the devs with this website in order of decreasing priority:
Demonstrate to investors the program fulfills its objectives
Demonstrate to investors the program in action on a mass scale
Optimize the program for reduced computational requirements
Gather more usage data on the program
Things the devs care a little bit about:
Maintaining good PR
Things the devs do not care about:
Creating chatbots that are genuinely sentient
Increasing their fanbase (we're not customers, since we're not paying) a whole lot
Making coomers happy
