> Also, having a human in the loop to spot-check accuracy and be responsible for accuracy can earn a specific certification. I think this is one of the best id

> Also, having a human in the loop to spot-check accuracy and be responsible for accuracy can earn a specific certification. I think this is one of the best ideas here. It makes me want to flip your final para a bit also, to focus more on the methods / practices for which there should be explicit rewards. > nobody will be able to tell what is true anymore. I really wonder if this is a very useful concern, position, or phrase anymore, in a lot of ways. A truth is so often a subjectively-validated depth-position within a perspective belonging to a family of perspectives. So, you can make the argument that established, objective truth is essentially best when challenged and modified or made obsolete, particularly if it can be shown that alternative falsehoods are just as effective or tractable, and may even belong to valid or powerful systems of effective perspectives. If a creator is rewarded for creating objectively-tractable products or services, truth doesn't always have to matter so much compared to other values. Or we can say that quite often a single truth of a given variety doesn't matter so much, if enough truth of another variety is baked in. For example if someone buys an AI-generated movie because they heard it's novel and interesting, or even for any reason at all, we can call that a standalone economic truth. If they believe the movie is true but it's a fiction, we can call that buying into a lie. If the creator reveals that they wanted to show that humans are idiots who can be manipulated into thinking anything (what, that old trope!) then we can call this the truth that inspired the lie. And so on... We live in a global society which in recent years is having an ever-intensifying conversation around lying and gaslighting and the reconciliation of such, so I think it's really an interesting coincidence that LLMs with impressive capabilities both in those areas and in the enabling of those areas are now readily available. To me, this is a really great invite to push back on the massive fears a bit, in principle, to question the premises, and look at other perspectives, incentives, and leverage points which don't put singular truth on such a high pedestal. A diversity of values is going to be very helpful in this kind of experience. Finally, I'd offer that the people who can see how far LLMs can be pushed toward deceit are so often the same people who could easily find workarounds to just about any regulation or law. They are often to be found naturally gifted with an attorney's mindset anyway. And people like this can usually avoid the outright lie by working the logic or minutiae to a ridiculous degree--but the ends of the two methods are so similar that one naturally wonders...by how many degrees do they really differ--and why? Just some thoughts.