Banning AI is silly
Good writing is good, bad writing is bad
It feels cleaner to move the blame to the tools, it is not
Each identity (human, bot, or org) develops it's own reputation
Mute & block those who do not meet your bar (personal or org)
(It does not matter if this post was written with AI)
Went to a russian sauna last night
It had variable cash discounts based on the food
Never seen anything like this
Surcharge for pea soup 8.8%; greek salad is 4.0%
Slobo retweeted
TLDR for crypto people:
> Be me
> Ran ETH validator while I lived on Wall Street
> Get call from federal prosecutor 3 years later
> HIM: "We're prosecuting hacker of Uranium Finance"
> HIM: "Need you to come testify that you ran a node for proof of Venue, its a law that requires crimes are tried where they are committed"
> Law doesn't make sense when applied to crypto, but whatever
> Me: "ok, how did you find me?"
> HIM: "Matching Etherscan and Verizon IP records"
> Fly to NYC
> Take oath and tesify
> HIM: "Cool thx bro"
> Post story on X
> Avi Eisenberg replies
> Avi's convictions were thrown out, not because he didn't do it, but because they couldn't prove it happened in the district he was tried
> Realize I'm the fix
> Realize everyone who runs a node is a witness to all crimes that occur on that blockchain
State of modern crypto law
AI Inbox for gmail launched and it isn't very good.
1. Has no context to action already performed
2. Turns succesful spam into TODOs
Maybe I'm spoiled by the high quality of dots & grok, but this is stone age
Official blog post up on Jev
typesafe.ai/blog
New Type: "Box"
Bounding Box: First type going beyond decisions
- Only takes instruction on what to find
- Returns confidence as well














