https://twitter.com/jinglingcookies Hey everyone, I love to research into various crypto verticals - anything and everything fun I’m from the Monad ecosystem team - if you are building something cool, do reach out! 🟣
Some thoughts around how to optimize UI/UX for users
https://x.com/jinglingcookies/status/1934191369947680818
Polldotfun
Wonder what does vote to settle mean
https://x.com/polldotfun/status/1932800297325506784
God is hungry for Context: First thoughts on o3 pro
https://www.latent.space/p/o3-pro
Try Your Best
Get rewards for posting about your favorite products
https://x.com/TYBXYZ
Noice: Making Cents of InfoFi
https://x.com/yeak__/status/1932871119951806556
MegaETH Ecosystem
• Euphoria: Mobile first derivatives app
• Valhalla: Perp and spot DEX
• GuessBest: Tinder style Polymarket
• Hop: Free and fast VPN
https://x.com/blockworksres/status/1924516164283355285?s=46&t=ywALktyM5ZCyt-zigN5Big
Agent Bricks: System to create agents optimized for certain tasks
With evals fine tuned to the tasks to constantly improve the agent over time
Seems like agents will be a lot more efficient with this
https://x.com/matei_zaharia/status/1932844943745446326?s=46&t=ywALktyM5ZCyt-zigN5Big
Stripe Acquires Privy: Building the Onchain Banking Stack
https://x.com/chuk_xyz/status/1933892516086002043?s=46&t=ywALktyM5ZCyt-zigN5Big
Yupp AI: A platform where users get to compare and get the best answers across LLMs
Raised $33 mil seed led by a16z
https://x.com/pankaj/status/1933625251851874568
ZeroMatter: Building a simulation environment for robotics training
Started by ex-Tesla engineers
https://x.com/jinglingcookies/status/1933814251275317693
NoLoCo by Gensyn AI
Training large data models over heterogenous gossip networks
https://x.com/gensynai/status/1933566620125048876
Customers are pulling away from Scale AI after the acquisition by Meta - in fear of their data being exposed to Meta
https://x.com/Jukanlosreve/status/1933754935101042929
Black Mirror for X
Reputation score based on how much content is AI generated
https://x.com/robj3d3/status/1933643641295417583?s=46&t=ywALktyM5ZCyt-zigN5Big
Fernwah: Messaging app that works without internet
https://x.com/offlineprotocol/status/1933561405262700631?s=46&t=ywALktyM5ZCyt-zigN5Big
This is interesting
I never noticed the thing about the desire for less human interaction when it comes to tasks where we want speed
How Are Consumers Using Generative AI?
https://a16z.com/how-are-consumers-using-generative-ai/
Came to think about offline context, things like who you meet, what you eat
https://x.com/jinglingcookies/status/1934074664185450979
Curious to see how this turns out
There has been many of such experiments but nothing that really stuck around, or managed to achieve venture scale in terms of revenue, user growth etc.
From the few snippets I see from TYB's X, it seems like they put a lot of effort into gamifying it, with things like obsessions (collecting 'gems' to populate their board)
Not entirely sure how the business model is, but feels like the fundamental bottleneck is how many merchants can be onboarded to provide incentives for users to post about them
And also over the long term, how does the posting contribute to revenue being made by the merchants
Staying Sober on Stables
What the stablecoin stack looks like, and how to emerge as king
https://x.com/nickjmatt/status/1932854901811609899
A Deep Dive on Galaxy Digital's Flagship AI Mining Center Helios
Source
Thru
Seems to be a L1 focused on allowing complex transactions by having bigger blocks
https://www.thru.org/
State of AI Agents
Worth looking into
• Ambient agents
• Agent UX
• RL for non-verifiable tasks
• MCP
• Agent memory
http://rlancemartin.github.io/2025/06/10/aie/
Some thoughts around Yupp AI
I think it's cool, something similar to LMArena where you get to rank the various LLMs used to provide answers
There's a couple of concerns
1. Paradox of Choice: This might potentially contribute to fatigue if users are required to look through 2 different set of answers. Remember, a large proportion of users right now probably delegate most of their decision making to AI, it's meant to facilitate and makes this as frictionless as possible
2. Precision: I usually still do some cross referencing to make sure what I'm getting from the AI is right. But if i'm just getting presented answers by 2 AI, it doesn't really help me in validating the answers
That being said, it's a pretty good way to 'rank' LLMs based on human preference, instead of ranking based on evals, which are typically not as relatable
https://x.com/pankaj/status/1933625251851874568
There's some pretty interesting reads in the comments
https://x.com/hosseeb/status/1933580930268168482
Anthropic: How we built our multi-agent research system
https://www.anthropic.com/engineering/built-multi-agent-research-system
Various forms of zkTLS
https://x.com/0xSerBito/status/1933529031363686735
Understanding 246 Club
Arbitraging between the lending markets and isolated pools
https://x.com/WinterSoldierxz/status/1933129402810020255
Are there any other agentic browsers available
https://x.com/jinglingcookies/status/1933760944892751945?s=46&t=ywALktyM5ZCyt-zigN5Big
https://x.com/fredaduan/status/1933738665571938550?s=46&t=ywALktyM5ZCyt-zigN5Big
Читать полностью…Why We Want Robots at Work but Humans in Art
https://docs.google.com/document/d/1-3q70tJV7A-Y99uGO6bd8Hrx4uE0iZtn-QrIa6wWag8/mobilebasic