LLMs use a lot of RAM as a fundamental part of their operation. The RAM is used to achieve the goal as efficiently as we know how. Even if you disagree with the goal needing to be achieved at all, the RAM usage is about as efficient as we can design.
Regular modern applications use a lot of RAM as an incidental or accidental part of their operation. Even if you think the tasks that they're achieving are of extreme need, the RAM use is excessive.
These problems are apples and oranges. You can hate both, or one, or neither. I know plenty of people who are in each one of those camps.
If you don’t think Chrome could be way more RAM efficient, and especially if you don’t think the things running inside Chrome could be more efficient, I have a bridge to sell you.
If you think acknowledging that fact (and the fact that there’s really not a great way to make LLMs more efficient) is “apologetics”, I cannot engage with you in good faith.
Regular modern applications use a lot of RAM as an incidental or accidental part of their operation. Even if you think the tasks that they're achieving are of extreme need, the RAM use is excessive.
These problems are apples and oranges. You can hate both, or one, or neither. I know plenty of people who are in each one of those camps.