If this is really where the researchers think these tools are headed (and I don't really doubt you on that point), then this is incredibly dangerous stuff. No m

If this is really where the researchers think these tools are headed (and I don't really doubt you on that point), then this is incredibly dangerous stuff. No matter how good your system is, the impact of implicit, unintentional, and non-targeted bias is huge on the sorts of content these systems will produce. But expose it to the levels of intentional manipulation present on the Internet of today, and these models don't stand a chance of producing something that safely does what you claim. I'm happy to be wrong about this, but I'm not seeing any discussion about the safety and security of using these systems. And if it's not even being discussed, we can be sure nothing's actually being done about it. Selling promises of active-agent computers interpreting human intent and summarizing information from the Internet without addressing this concern is irresponsible at this point.