jun_lung

Because why should you have to load "here's how to update this skill!" information into the context window every time you use it? Would you expect the agent to go to that URL and look for more recent skill files every time you use it? Is this a real question?

I think it helps to draw a clear boundary between what is "content" and what is "art", since they ultimately serve different purposes and have vastly different factors that go into their creation. To me, "content" is anything created with the goal of maximizing consumption, whereas "art" is an attempt to express or give form to some kind of idea or emotion. The lines blur when you get to things like film, where those two goals often compete, but it's pretty easy to put things into these two categories.

And this isn't to say that "content creation" is any less valid than "creating art". There are plenty of "content creators" who know exactly what they are making at the end of the day (a product) and don't hide that fact or try to dress it up under the pretense of being anything more.

Reminds me a lot of the book Manufacturing Consent, where it breaks down the split between 'journalism' and 'propaganda'. As long as there is a monetary incentive to emphasize certain things and ignore others, it will always ultimately be a form of propaganda.

Not to continue piling on here but if we want to shift from a human consumer to an LLM consumer, I think the WORST thing you can do is obscure raw data behind some kind of "widget"

Personally I think we SHOULD keep returning data when we are using a protocol built specifically to transport data. I am struggling to see a reason for this to exist, and it almost feels like someone saying "I was sick of my REST API returning a bunch of boring old JSON!"

For auth, how does this handle things like 2FA or cases where, for first login from a new IP/whatever, there's an extra verification step (email, etc)