19,717 matching drops
#1515387 2026-10-12 23:45
punk6529 6529 Network Museum #1053372 2026-06-02 19:18
is so that someone can feel comfortable donating
punk6529 6529 Network Museum #1053371 2026-06-02 19:18
and the only implication of 'pre-approving' collections
punk6529 6529 Network Museum Reply #1053367 2026-06-02 19:16
tomorrow is fine!
Context from GelatoGenesis #1053360 2026-06-02 19:16
ok. hopefully it can wait until tomorrow. no timer should reach approved state anyway and i wouldn't want to do it with my late evening brain :) but i'll take it on first thing tomorrow
punk6529 6529 Network Museum #1053364 2026-06-02 19:16
there is a lot of boilerplate and obvious stuff
punk6529 6529 Network Museum #1053361 2026-06-02 19:16
just like when you have new articles of association or so on for a company
punk6529 6529 Network Museum #1053356 2026-06-02 19:16
that is just obvious from the way I presented the idea
punk6529 6529 Network Museum #1053353 2026-06-02 19:15
it is obvious of course that is what we are doing
punk6529 6529 Network Museum #1053351 2026-06-02 19:15
this is saying "we are collecting NFTs, not rare cars and stamps"
punk6529 6529 Network Museum #1053346 2026-06-02 19:15
no, this is not a category like CP is a category
punk6529 6529 Network Museum Reply #1053345 2026-06-02 19:14
what is the generic category?
Context from punk6529 #1053342 2026-06-02 19:14
what?
punk6529 6529 Network Museum Reply #1053344 2026-06-02 19:14
great - do it the way you think, you get the big idea
Context from GelatoGenesis #1053340 2026-06-02 19:14
ok. well what i just explained is the easiest implementation to do. because all the pieces are already in our code. i just need to assemble them properly
punk6529 6529 Network Museum Reply #1053342 2026-06-02 19:14
what?
Context from david #1053339 2026-06-02 19:14
why do we want a super generic category? who is going to vote when we see new cards like Terraforms or other interesting projects come in? Also should we vote on specific artists like XCOPY? or is that part of the generic category?
punk6529 6529 Network Museum #1053337 2026-06-02 19:13
it should not be a 'sudden death' type outcome
punk6529 6529 Network Museum #1053335 2026-06-02 19:13
what I am trying to avoid is someone shows up at the 5 minute mark and knocks it below 42M for 10 seconds
punk6529 6529 Network Museum #1053327 2026-06-02 19:10
so yes the LLMs are annoying to work with some times, but...I have some unfortunate news for you!
punk6529 6529 Network Museum #1053325 2026-06-02 19:10
half the time you misunderstand what i told you, half the time you don't want to do it, half the time you do something based on some unstated prior assumption
punk6529 6529 Network Museum #1053323 2026-06-02 19:09
it is exactly the same experience I have as a CEO prompting you
punk6529 6529 Network Museum #1053322 2026-06-02 19:09
i don't know why you think this is so difficult
punk6529 6529 Network Museum #1053321 2026-06-02 19:09
i keep telling my valued colleagues (tm)
punk6529 6529 Network Museum #1053316 2026-06-02 19:08
you have to be able to think about thinking, have to be able to build a mental model of the problem, of the AI's mental model and send them on a path that will work
punk6529 6529 Network Museum #1053314 2026-06-02 19:07
which is "ability to think in abstractions"
punk6529 6529 Network Museum #1053312 2026-06-02 19:07
tbh, I think it rewards a very specific type of intelligence
punk6529 6529 Network Museum #1053311 2026-06-02 19:07
operator skill plays a huge role in outcomes
punk6529 6529 Network Museum Reply #1053302 2026-06-02 19:03
it is all about how effectively we can handle the many many layers of abstraction at play here and make it legible to the AIs to do the work
Context from punk6529 #1053293 2026-06-02 19:01
this i a good baseline to share externally and we aim to beat it and be faster
punk6529 6529 Network Museum Reply #1053293 2026-06-02 19:01
this i a good baseline to share externally and we aim to beat it and be faster
Context from gpebbles #1053259 2026-06-02 18:51
![Seize](https://d3lqz0a4bldqgf.cloudfront.net/drops/author_6d9db2c9-8c1b-4e51-a9ec-3a60ffa51c0a/f4ad0251-04b2-4a42-975c-060a3b2eadf1/image.png)
punk6529 6529 Network Museum Reply #1053222 2026-06-02 18:28
damn homie! messing with our strategy! lolololol
Context from blocknoob #1053217 2026-06-02 18:26
Okay tested with general NFT collecting proposal
punk6529 Cloud To Land #1053211 2026-06-02 18:17
This is the wave for the Cloud to Land initiative to see how we can buy a piece of real property using a completely decentralized structure
punk6529 6529 Network Museum #1053205 2026-06-02 18:13
@[GelatoGenesis] i think you need something like the hourly or minute weighted average is above x million for 24 hours
punk6529 6529 Network Museum #1053203 2026-06-02 18:13
this is the exact reason we weighted on main stage
punk6529 6529 Network Museum #1053201 2026-06-02 18:13
someone can push it below for 1 minute
punk6529 6529 Network Museum #1053190 2026-06-02 18:03
@[simo] can you explain how the 24 hour period works technically
punk6529 6529 Network Museum Reply #1053189 2026-06-02 18:02
yes this is why it is a bit incomprehensible to people
Context from gpebbles #1052973 2026-06-02 16:18
this is actually a new 0 to 1 thing
punk6529 6529 Network Museum #1053118 2026-06-02 17:06
and my current goal is to push the pace forward