Robo excellent (spacex80)

What do you make of this? No Farcaster-launched assets among the top traded tokens on Base, few among the top performers (if we compare established tokens). Farcaster is becoming a solid crypto social, miniapps are adding value & at investable mcaps - but user numbers, mindshare isn't here yet.

• 1 reply
• 4 recasts
• 7 reactions

llm's are non-deterministic. you can't predict with certainty what they will do time after time. 1 million + messages have been sent to @bankr not just by the team, but in public right on the timeline and through the private terminal, and other platforms we support like xmtp and farcaster. you cannot build an agent that handles money without rigorous testing of all different types of requests and scenarios. we don't just handle single requests we handle entire conversations with multiple participants. we've been building in public since the end of last year. we have a ravenous community of believers behind us helping us build the best transactional crypto agent day in and day out. this is the bankr moat. this is why we win. its $bnkr bish.

• 1 reply
• 4 recasts
• 7 reactions

#ETH good to the mars🚀🚀🚀

• 1 reply
• 4 recasts
• 6 reactions

I’d be curious to see a data-backed analysis of the average engagement breakdown for casts that have, say, at least 500 views (as an arbitrary threshold for visibility — could be tweaked later). My very empirical observation so far on my own is that for every hundred viewers, ten will like, three will reply, and two will recast (with some overlap between them). So, ~90% merely consume content but don’t interact with it in any way. I’d love to see what kind of distribution this empirical law, if any, follows around the mean. And how notorious IRL users possibly skew that trend by causing artificial replies

• 1 reply
• 6 recasts
• 7 reactions