[cosmic_cheese]

At the risk of becoming the infamous iPod and Dropbox posters, I really don't think so. My browser having an LLM directly integrated adds nothing for my use cases that couldn't be accomplished with a web service or dedicated tool/app. For me, an integrated LLM running concurrently with my browser just represents a whole lot of compute and/or network calls with little added value and I don't think that this is unusual.

[zamadatix]

Better yet, if an LLM does add value to the use cases why is it that I have one "integrated" LLM when editing a document in the webpage, another "integrated" LLM in the browser, and then an "integrated" LLM in the OS. If there is value to be had I want it to integrate with the different things on the system as they exist just like I do, not be shoehorned into whatever company abc decided to bundle with just their product(s) too.

[cosmic_cheese]

Yep. I mention this in my other reply, but having the LLM be system-level (and preferably, user replaceable) and leveraged as needed by applications (and thus, not redundant) is clearly the best model. Apple is currently the closest to this, offering system level third party LLM integrations, but a Linux distribution would be the best positioned to achieve that goal to its fullest extent.

[brians]

Having something that read everything I read and could talk with me about it, help remember things and synthesize? That’s awesome. Follow links and check references.

[cosmic_cheese]

This use case feels better served by a dedicated utility with a specialized UI rather than shoehorned into a browser. It'd fit the macOS services model (which adds items to context and application menus, e.g. "Research this…" when right-clicking a link or text selection) and could optionally also be summoned by the system app launcher (like Spotlight).