Logan Kilpatrick

Like you could do some hacks to sort of

get slightly farther.

And like, we did show a bunch of like research, um, of what it would look like to bring 10 million to people.

And even with the original Gemini launch showed some of that in, in like practice and production environments, it becomes extremely like very, very, very costly.

Um, and like not easy to maintain and continue to scale up.

So I do think we'll need some like architectural, uh, innovation at the model level in order to enable things like a hundred million tokens.

um, which I'm excited about and I think the world needs.

So I'm, I'm hopeful we'll keep pushing the rock up the hill.

What's a hundred million token use case.

Yeah.

I mean, some of these code bases is actually a good example of like, if you look at like a large company and if you're using a mono repo, really interesting to see, like, uh,

like you probably maybe a hundred million tokens is like too much or is like slightly on the extreme of this, but like accumulated through your lifetime, you actually do have a lot of this data.

I think the challenge then becomes like, how do you, and like the attention mechanism in language models and transformers specifically doesn't have this intrinsically in it, but like, how do you up sample the right data and down sample the wrong data, all that stuff.

So I think you'll need,

These sit like human memory has this like interesting has all these interesting mechanisms to make sure like the stuff that's not useful is removed from your memory and the stuff that is useful cemented and then you build these this sort of abstraction layer on top of.

um like core ideas that you've had i assume we'll get something closer to like what looks like memory um because then it it decreases like you don't need 100 million you probably need like 10 million with the system on top of it um and that's much more reasonable than 100 million totally that makes sense

Yeah.

This depends probably a little bit on like the use case.

You could definitely do something like that.

Um, I would, I would suggest folks to like, try to do some validation as far as like how accurate that is.