Lei Yang

priority allocation when you have real-time trading.

Yeah, the transparency part is really important because I think as an independent trader, I think our mechanism could actually empower more independent, curious traders because as an independent trader, you definitely don't want to navigate the maze of which...

Right, so it's quite interesting.

I think for, I think to help agents, you need the block space to be sufficiently cheap so that they can try and error.

I think something that's very interesting that agents have immense amount of energy, I mean, unlimited amount of energy to do that humans do not, are just try and errors.

Like, I get angry if MetaMask does not work on megates for like after two tries, I just throw my phone out of the window.

Oh, shit.

Agents don't care exactly, as you mentioned.

So I think being cheap enough for agents to work with.

Second, some inherent features.

So one feature we have been talking about, we have been thinking about is, one, you can just imagine this user flow, right?

A user wants to achieve something, right?

Hey, I want to swap...

a thousand US dollars for what 0.5 ETH I hope it's not 0.5 but seems that it's the case but so I want to I want to swap a thousand dollars for 0.5 ETH okay and just agent figure out like what app just get me 500 dollars sorry get me 0.5 ETH right

So now the agent will kind of try to use different paths, different DAXs, different aggregators.

So it's almost like intent protocols.

I think with agent, you almost have your own intent solver, right?

So then the thing is, as a user, I would worry, hey, what if the agent just goes crazy, right?

There's always the possibility.

I think the thing with any like human in the loop, but now agent in the loop system is that there's always the possibility of just random errors.