Zishuo Zhao @w3w_zhao

Maybe we've never trained AI. Its the AI that trained us. Since the first kindling of fire... Till we discover, rather than invent, it.

For incentive design in the decentralized world... Is peer prediction all you need? I don't know... But it is indeed a tempting blue sea.

Success is uncertain, but entertainment is guaranteed! ✨

If you manage to break impossibility results that are based on certain assumptions, it is amazing work. But you don't need to shout all over the Earth that previous results are "wrong". Einstein did not say Newton was wrong. He proposed his new theory that improves Newton's.

David Lancashire 🟥 @dlancashi

8 months ago

If you manage to break impossibility results that are based on certain assumptions, it is amazing work. But you don't need to shout all over the Earth that previous results are "wrong". Einstein did not say Newton was wrong. He proposed his new theory that improves Newton's.

2 0 6 607 1

Please avoid investing on the project Saito. I have never seen a serious researcher frequently claiming that his work is "the best in the world" or "the only one that..." and recklessly claiming others' works are "wrong". I believe that it is actually a scam.

HD Selecta @HD_Selecta

8 months ago

Please avoid investing on the project Saito. I have never seen a serious researcher frequently claiming that his work is "the best in the world" or "the only one that..." and recklessly claiming others' works are "wrong". I believe that it is actually a scam.

1 14 35 3K 2

And I feel the security of a system might depend more on the the systematic design rather than users' behavior. If the system is vulnerable, its public security is inherently downgraded even if nobody exploits. So in my opinion the security is not affected by the users' actions.

David Lancashire 🟥 @dlancashi

8 months ago

And I feel the security of a system might depend more on the the systematic design rather than users' behavior. If the system is vulnerable, its public security is inherently downgraded even if nobody exploits. So in my opinion the security is not affected by the users' actions.

0 0 0 399 0

Download Image

I think we might be using different models. In my model I assume that the users are selfish and they only care about their own utility (valuation of the tx confirmed minus the tx fee payment).

David Lancashire 🟥 @dlancashi

8 months ago

I think we might be using different models. In my model I assume that the users are selfish and they only care about their own utility (valuation of the tx confirmed minus the tx fee payment).

0 0 0 399 0

Download Image

Thanks for your comments. Actually it is an early work of mine as the reviewing time of OR is quite long, so the modeling might seem a little outdated in 2025. It seems that you are using a quite different model. Could you elaborate on the routing work and your detailed modeling?

David Lancashire 🟥 @dlancashi

8 months ago

Thanks for your comments. Actually it is an early work of mine as the reviewing time of OR is quite long, so the modeling might seem a little outdated in 2025. It seems that you are using a quite different model. Could you elaborate on the routing work and your detailed modeling?

2 0 6 607 1

Paper accepted by Operations Research! [2209.13099] Bayesian Mechanism Design for Blockchain Transaction Fee Allocation

#EldenRingDLC Carved words coalesce. "I abandoned here my Verifier's Dilemma."

"如果有一天，AI真的统治了世界，当它想起人类中那些最聪明、最富有的那一部分人，曾为它的面世而如此奋不顾身的时候，它或许也会善待我们的吧。"

My recent work on Proof-of-Learning arxiv.org/abs/2404.09005 (A new version with substantially improved theoretical results is coming!)

Old account was hacked and suspended... This is me!