arya (fka. arjun) 🏴

The recording of my talk from @bankless Summit is up! This unconventional talk dives into the WHY of crypto, something we unf don't think about enough anymore. I hope this is helpful context for folks who are new to the space, and a galvanizing reminder for the veterans. 🏴

@0xJoxes Thanks lol. It's a pity it got bogged down in the misinformation that was being created at the time by projects that wanted to lock users into their own ecosystems. I was also unf too burnt out and not structurally incentivized to try to fight it back then.

xERC20 (ERC-7281) was great for its time, but unfortunately underestimated. The relevance of it in this opportunity was the establishment of rate limits, as an honest step to address the real challenges the ecosystem had (and still needs) to face, such as: - Any verification method is inherently risky by nature, the only thing that changes is the axis observed - Monitoring processes lacking public accountability and will fail at some moment - The need for clearer ways to express risk aversion @arjunbhuptani was really a visionary in this design, and in the crosschain vertical in general. ethereum-magicians.org/t/erc-7281-sov… github.com/defi-wonderlan…

Ok - this is incredibly sick. Private, p2p inference running on idle GPUs could legitimately save us all from technofeudalism if it works. Just spun up a darkbloom node to learn more.

Back to X after a long break. D̷̗̐ḭ̴̾ḋ̴͖ ̵̪͌y̷̲̿ó̸̠ů̸̫ ̶̝̈m̶͍̂i̷͚͒s̸͚̚s̷̥̈ ̴͇̚m̸͚̈ḛ̷̋?̷̱̐

Not sure I have ever been as productive in my life as in the last ~10 weeks: An extremely challenging yet worthwhile goal. The crazy progress of AI coding and heavy use of it. A very small team, world-class skills, and now the ability to multiply output thanks to AI. The result is @etheconomiczone — with a live devnet already: eez.dev. What is this? It will be: a) A new rollup that is fully composable with Ethereum. This means you can do either cheap transactions on this chain that only interact with contracts on this rollup, or more expensive transactions (similar to L1 costs) that can interact in a fully composable way between L1 and L2. So, for example, if you trade on this rollup, the routing can dynamically decide to use only L2 liquidity for a small trade, or for a larger trade — where the extra gas costs are worth it — to use both L2 and L1 liquidity. b) A whole framework that will allow new rollups, as well as existing L2s and sidechains (yes @gnosischain!!) to integrate into this “(Ethereum) economic zone.” Imagine: one could make a trade today simultaneously using Ethereum liquidity, but also Arbitrum, Base, or all the other L2s. This is what EEZ will allow. Now, let’s talk about the tech: Essentially, there are two core concepts that make all of this possible: 1) Proxy contracts 2) Real-time proving 1) Proxy contracts are basically a way to overcome the challenge that ~99% of all contracts are written in a way that only deals with addresses from a single chain. For example, a token or an NFT can be sent to an address, but not to an address on a specific other chain. Proxy contracts fix this. An address “A” on chain “n” will get a deterministic proxy representation, “A*”, on all other chains. So now, if, for example, A is a DAO on Ethereum and it should control, say, a fee switch in contract D (a DEX) on another chain, this can easily be done by setting A* as the owner of D. A (the DAO), on the other hand, can now call D* (the proxy representation of the DEX contract D) on Ethereum. All the cross-chain message passing in between is abstracted away — the contracts just call another contract and do not realize it is actually on another chain. 2) Real-time proving The proxy design already addresses the problem that there is no widely supported cross-chain message-passing standard in EVM land. So it alone would already be helpful for asynchronous calls, or better, calls that do not expect or require a return value. However, this would still not quite bring us to “synchronous composability.” Imagine a DEX trade: you do the first part of the trade on L1 and the second part on L2. You want to know the result of the second part — and if it did not get you the expected amount, you want to be able to revert the first part. This requires the call that triggers the DEX trade on L2 to have a return value. This was long assumed to be impossible - but with real-time proving, it no longer is. Basically, because Ethereum blocks are only produced every 12 seconds (and this would still work with, say, 6 seconds), that is now long enough to build and prove the L2 block that contains, for example, this L1→L2 DEX trade. If you want to learn more, don’t miss the talk by @tw_tter and @jbaylina on Tuesday at @EthCC! Let’s make Ethereum ONE again!

@tayvano_ lmfao

take me back to a better time

@gakonst "hub based payment channels"

@0xwassie Appreciate it! 😄

@arjunbhuptani great move everclear, hope this cuts through the stablecoin noise and actually delivers some real liquidity

Everclear just shipped one of our most requested features by users. In effect, this turns the protocol into crosschain Curve for institutional flows. Any stablecoin can be settled into stables on any other chain, fixing fragmentation between the 10000 diff stables out there.

A big step for the intent stack! 🥞 Everclear is rebalancing @sprinter_ux’s Stash — an onchain credit protocol giving solvers instant access to zero-collateral liquidity. More capital access → more solvers → more decentralized clearing. 👇

We have several huge pilot partners for going live with this feature in the coming weeks. Excited to showcase how @EverclearOrg can solve problems not only for bridges, but also in large scale DeFi and payments. ✨

Crosschain Asset Settlement is live! 🎉 Protocols and solvers can now instantly rebalance between similar assets like USDC⇄USDT within Everclear—seamlessly integrating swaps into clearing flows for even better netting efficiency. Crosschain just got a whooole lot smoother. 🧈

daira is an og, a principled proponent of human freedom. as opposed to the waves of crypto tourists, who have flooded the space with stuff like '99999x pump fun' or 'state endorsed hedgefund eurodollar'.

Sub-minute transfers are here! 💨 Priority Settlement gives solvers near-instant rebalancing that integrates into clearing flows—key for crosschain intent spikes when speed trumps cost. Moving beyond a one-size-fits-all model, we're paving new pathways for ecosystem adoption.

Our largest transaction to date recently cleared! ⛅ 1378 WETH (~$5.5M) closed in just 2hrs for only 5bps. Onward to even higher highs. 😎

Today, we’re upgrading Eco Routes to utilize @EverclearOrg as a primary partner for liquidity rebalancing. Now, cross-chain stableswap execution on Eco is faster and more cost-efficient at larger order sizes.

LLMs struggle with fragmented context and limited memory. Almost every organization I know is looking for solutions now. Here’s 37 products we explored recently that help build an org-wide AI OS 🧵 ---- 1/ AI OS / copilots @slashyai - very smooth workflows, no memory. getden.io - workflows, but feel slower. cofounder.co - great onboarding, layered, long‑running memory, but a bit slow. flowith.io - workflows with knowledge. @EragonAI - enterprise/business data focused, locked-in founder @joshsirota and notorious @cluely is also tackling the memory problem! apart from them, we just found everylab.ai, @thelobbyai, uOS, @steve_aios and acris.ai but haven’t tried yet! ---- 2/ General memory apps @Letta_AI - fantastic playground for building stateful agents, sleep-time compute concept, love it. @mem0ai - invisible chrome extension that captures your LLM inputs across the board and builds memory around that. supermemory.ai - similar to mem0. kinic.io - like mem0 but encrypted (zk/verifiable compute). @openInfer Mementos - local memory system. @zep_ai - open-source knowledge graph (19k Github stars). also we found @hyperspell, capi.dev, @Threadbaire, honcho.dev and myneutron.ai who we are going to try soon! ---- 3/ Meeting-focused AI-workspaces Well-known Granola.ai - my favorite note taking tool turning into AI-workspace. hypernote.com - on-device, privacy-first Granola. tana.inc - a combo of Notion, Motion and Granola. ---- 4/ Connectors @composio - solid MCP aggregator used by many of the apps above. @airweave_ai - massive knowledge ingestion and search platform with real-time sync. mcp-use.com - Vercel for MCP. ariata.com - open-source data ingestion, quite experimental. @PraxosInc - data ingestion triggered via SMS/WhatsApp/Telegram. ---- 5/ Memory-based apps @mymind - privacy-first, notes-based workspace. mem.ai - an AI notes app with deep search. nia.ai - context for coding, robust evals (kudos to 18yo @arlanr) @pallyai - CRM/intro-making focused app. Obviously there are many-many more but its what we tried including poke.com and twinmind.com ---- 6/ Github/Cursor-based workspaces CybOs by @sgershuni - personal Cursor-based AI workspace, very flexible, set-up takes time. Cursor super-agent by @sevaustinov - for teams, same pros and cons as CybOs. ---- After trying all, I spent most of my time with: Slashy, mem0, supermemory, Letta, Composio, airweave, and, of course, Granola. Cursor-based set-ups are fantastic and we spent a lot of time playing with them but it’s quite hard to get an org-wide adoption. If you are also looking to build an internal OS with memory - would love to chat and hear your learnings!

Chain Abstraction In Action! No user should NEED to know which chain they're on. Just like how most users don't care (or are even aware) whether an app is hosted on AWS or Azure: the same will go for Web3. Here’s how Everclear makes this a reality. 👇

Exciting Monthly Highlights! 🤍 From hitting $2B in total volume cleared to onboarding Eco and Sprinter for rebalancing—September was full of huge milestones. With infra upgrades + new upcoming revenue streams, Q4 will be packed! And this is how we’ll reach trillions…👇

🥅 What is Netting? After filling your intent, Solvers get paid back on the source chain—but often need to transfer funds back to keep fulfilling more. Everclear aims to make this FREE by matching opposing transactions to net them out, meaning no funds move at all!

@Tiger_Research_ 🫡