Crypto abstraction!
Thing's need to be explained better!
Instead of being manipulated for one's own financial gain!
So here's
A few handy tools I've come across to lower the barrier's to entry for newcomers, for a overall better user experience!
🧵👇
teapawt.net/growthepie_eth/status/…
J-pEgG🇬🇧ZKEVM🥷03/01/09Protocols are not app's👊 retweeted
If we are unaware of the potential before us, we will not achieve everything we are capable of at this moment. What we are building with our software and laws at this time is nothing less than a completely new social structure. It deserves all of our attention.
J-pEgG🇬🇧ZKEVM🥷03/01/09Protocols are not app's👊 retweeted
By the way, this could be relevant to cryptography and Ethereum.
For instance, in ZK prover networks: if agents follow a standard pattern and all choose the same prover, the others will sit idle though this is just a theory!
2/2
Efficiency of information utilization by AI agents.
Imagine you have an excellent book, but if you read it inattentively or get distracted by seeing someone else with another good book-you won't derive any real benefit. This research paper examines a scenario involving a group of AI agents: why is it that while individual agents may make sound decisions, the collective outcome ends up being poor?
The Experiment:
16 agents, each with a specific task.
2 servers with limited bandwidth.
The first server goes offline temporarily, then comes back online.
A shared status report, delayed in its delivery, indicates that the first server is unavailable.
The study tested how agents react to outdated information and the actions of other agents.
* See the image for the specific models used.
Findings:
1) Information regarding a system's recovery alters agent behavior.
2) Shared evidence of success helps agents (a "celebration" boosts morale, haha).
3) Identical information can lead to inefficient coordination (for example, if agents see that the first server was working successfully, they might all flock to it simultaneously, leaving the second server underutilized).
4) Diverse input data helped optimize the workload.
This provides a deeper understanding for evaluating and studying how agents operate.
1/2
arxiv.org/html/2611.22632v1
J-pEgG🇬🇧ZKEVM🥷03/01/09Protocols are not app's👊 retweeted
No organisation using Aragon was unable to access their funds or upgrade their protocol yesterday.
Decentralised, censorship resistant, permissionless. Unstoppable.
J-pEgG🇬🇧ZKEVM🥷03/01/09Protocols are not app's👊 retweeted
Where is the tech for tribes?
We're on the me-me phase of the cycle and that only leads to incumbent tribes (corp) to max extract from the individual.
There has to be a mindest change where AI is at the middle of a group of people not as a boundary between them otherwise we can't get collective bargaining power.
👀 @bonfiresai modular sensemaking knowledge graph structures ✍️🤓
as life becomes more complex for all of us, we need tools that will help us navigate it and thrive above it. it's genuinely a big gap in consumer AI and tech these days. I want to live more, not less. more reality, less screens. information processing tools ftw
1
24
🧐@bonfiresai is being built to fix everything highlighted in this essay.✍️
@jb87ua @at0x_eth
coordination debt measured in to value in a useful way!
Sensemaking knowledge graph first map, that is a modular self updating information network structure not a final framework.
What actually breaks when a team struggles to make things happen?
This new essay explores how challenges change as organizations grow, from founding teams to large, complex systems.
alexsotodigital.substack.com…
1
4
84
J-pEgG🇬🇧ZKEVM🥷03/01/09Protocols are not app's👊 retweeted
Six kinds of wealth the world has learned to name.
The seventh was under their boots the whole time.
Not the new oil.
The old ground.
sync.soulbis.com/p/the-seven…
@audreyt inspired by the soil metaphor
🤖 Made with AI
J-pEgG🇬🇧ZKEVM🥷03/01/09Protocols are not app's👊 retweeted
If you're in crypto or AI pivot to knowledge supply chain logistics
J-pEgG🇬🇧ZKEVM🥷03/01/09Protocols are not app's👊 retweeted
Step 3: Collision Resistance. Since there is a vast amount of potential input data but the hash length is limited, different files will inevitably share the same hash (this is not a constant occurrence, nor is it as dire as it sounds). This poses a greater risk than the preimage scenario; therefore, hash functions must distribute their outputs chaotically to ensure that identical matches do not occur.
Now that you have learned the recipe, go forth - create, and protect your information!
A "Kitchen" for Preparing Hash Functions: Key Mathematical Challenges for Maintaining a High-Quality Hash Function
* A hash function is the transformation of a data array into a short string of fixed length.
Imagine you are a chef in a "hash kitchen," preparing high-quality, well-crafted hash functions.
Let’s explore the recipe and the preparation process!💫
Step 1: Preimage Resistance. Your hash must possess the property of "one-wayness" it must destroy information about the original data structure. (Incidentally, to reverse this process and recover the original data, one would need to make 2^255attempts a scale so vast that it is practically impossible, as is the case with SHA-256.)
Step 2: Second Preimage Resistance. You must craft your hash function so that changing even a single component of the input causes roughly half of the hash's bits to change in a random fashion. This ensures security, guaranteeing that no one can surreptitiously alter the data (which is "signed" by the hash function). In other words, it is impossible to modify the information while keeping the hash identical.
J-pEgG🇬🇧ZKEVM🥷03/01/09Protocols are not app's👊 retweeted
A "Kitchen" for Preparing Hash Functions: Key Mathematical Challenges for Maintaining a High-Quality Hash Function
* A hash function is the transformation of a data array into a short string of fixed length.
Imagine you are a chef in a "hash kitchen," preparing high-quality, well-crafted hash functions.
Let’s explore the recipe and the preparation process!💫
Step 1: Preimage Resistance. Your hash must possess the property of "one-wayness" it must destroy information about the original data structure. (Incidentally, to reverse this process and recover the original data, one would need to make 2^255attempts a scale so vast that it is practically impossible, as is the case with SHA-256.)
Step 2: Second Preimage Resistance. You must craft your hash function so that changing even a single component of the input causes roughly half of the hash's bits to change in a random fashion. This ensures security, guaranteeing that no one can surreptitiously alter the data (which is "signed" by the hash function). In other words, it is impossible to modify the information while keeping the hash identical.
J-pEgG🇬🇧ZKEVM🥷03/01/09Protocols are not app's👊 retweeted
Really sad to see what’s happening in Paraguay, where over 60% of the population is losing access to their funds held in banks.
As DeFi and stablecoins gain adoption among fintechs, we must remember the bigger mission, which is building a self-sovereign financial system where access to your own money is a fundamental right, not a privilege granted by an intermediary.
BREAKING: 🇵🇾 What just happened in PARAGUAY is a WARNING for everyone on earth.
More than 3 million people (60% of Paraguay's adult population) have lost access to EVERYTHING. Their bank accounts. Their pensions. Frozen. Inaccessible.
Paraguay’s largest digital bank, Ueno Bank, has collapsed.
The president of the Central Bank (who is deeply connected to Ueno) has just RESIGNED. Eight high-level executives including the founder and the Bank President are being detained.
The people have taken to the streets to demand their money back. And rightly so.
This is the model Milei and the rest of the Epstein class are pushing. This could happen ANYWHERE. Best to save in physical gold and physical silver, imo.
J-pEgG🇬🇧ZKEVM🥷03/01/09Protocols are not app's👊 retweeted
trust is not a score,
it is a keeping,
an orbit declared,
a leaf the trees hold,
to one day leave the forest,
for a breath of fresh air,
find a neighbours edge,
and root again.
J-pEgG🇬🇧ZKEVM🥷03/01/09Protocols are not app's👊 retweeted
just another mage sharing their spellbook 🧙
Awesome to be mentioned here on the open frontier of cryptanalysis.
SHA-256 is not broken, but its security margin lost 1 more round this year:
Last week I found a collision attack on reduced 39-round that improves prior art by over 30 bits and the first differential attack on that round better than generic search on @yukonresearch Hashsmash board.
The attack rests on the same method as prior art ePrint 2026/1120 with new insights that first 2 state words could remain free, allowing a critical condition at step 16 rebuilt as a claw-search problem rather than a probabilistic filter. Took days of watching SAT logs, burning cores and arguing with Claude on what's possible to get a result close to generic search.
Since my initial submission, other hashsmash participants sraiw, @cyberaxe, @privacymage, @lamb356_ and @geeknik pushed it down by over 6.5 bits, currently estimated at 118.15 bits. The tree of submissions is hard to follow, so I put the construction, insights and numbers into a blog and am working on detailed technical report.
It's fascinating to see an initial structural observation develop into an improvement in the reduced-round collision frontier within days. AI is now a real aid in cryptanalysis and open collaboration speeds it up.
J-pEgG🇬🇧ZKEVM🥷03/01/09Protocols are not app's👊 retweeted
Ethereum Network Development and Updates for 2026
Let’s look at some of the updates:
Lower transaction costs (see table)
Reasons include increased blob throughput, PeerDAS (selective blob verification by validators instead of downloading every blob), and raised L1 gas limits.
Next, improved UX, including native account abstraction; for example, EIP-8141 builds smart contract logic directly into the protocol—an important step toward post-quantum readiness.
"Glamsterdam" proposals:
EIP-7732 and EIP-7928
7732 – formalizes the separation between the validator and the builder ; this eliminates the need for centralized relays and removes blind spots during block creation.
7928 – creation of block-level access lists.
Ethereum is evolving every day.
ethereum.org/latest/building…
J-pEgG🇬🇧ZKEVM🥷03/01/09Protocols are not app's👊 retweeted
Learn more about each point and the relevant EIPs: ethereum.org/roadmap/privacy…
Ethereum Network Privacy.
We have Ethereum, and Ethereum has a roadmap; within that roadmap, we can highlight specific priorities:
Private reading: accessing information without revealing the data being retrieved or the specific details being viewed.
Private writes: submitting transactions that are uncensored and unlinkable to their point of origin.
Private proofs: using zero-knowledge proofs to verify identity or data without disclosing personal information.
J-pEgG🇬🇧ZKEVM🥷03/01/09Protocols are not app's👊 retweeted
Software development has become a lonely thing, and I don't think that's progress.
Everyone's a lone wolf now. Just you and your Claude agent, shipping your own little SaaS, burning tokens so the agent can remember what it did yesterday.
Moloch loves a lone wolf.
I think the more interesting outcomes come from swarms. People and agents building on the same memory, picking up where each other left off, adding up to something none of them could build alone. Some sort of Voltron.
That's what we've been building at @bonfiresai . Shared memory for humans and agents, organized without spending any LLM on it.
Hanging out infrastructure.
It's kind of my life's work. More soon.
J-pEgG🇬🇧ZKEVM🥷03/01/09Protocols are not app's👊 retweeted
Note: there's a subtle distinction between L1 <> L2 interop and Ethereum <> alt-L1s. The zk light client is afaik mostly used for the latter. In my view, it matters to increase distribution of Ethereum assets towards alt-L1s too, as securely as possible, and the light client is the right first step before fully zk-verified proofs of finality.
But L2s offer more than alt-L1s: With a stage 2 canonical bridge between them and L1, L1 assets moving to L2s have the same security guarantees they'd have if they were still on L1. Contrast with L1 assets moving to an alt-L1: should the alt-L1 be corrupted, and its validator set reports the wrong state of its domain, alt-L1 operators could steal of freeze bridged assets.
So for bridges to alt-L1s, verifying Ethereum's confirmation is the best you can do: it removes trust in a third party's report. But it can't remove trust in the alt-L1's own validators. Only L2s with stage 2 bridges carry L1 security end to end for bridged L1 assets.
Either way, improving Ethereum's settlement layer is key to extending the reach of its assets and services.
See also @decentrek 's tweet for more nuances teapawt.net/decentrek/status/20900…
It's ironic that Base bridges faster to Solana than to Ethereum
Right now, the Base -> Ethereum bridge operates in a "multi-prover" model, where if a TEE and a ZK prover agree on Base's state transitions, then a Base state update can settle on Ethereum in 1 day. Why 1 day? Because there might be bugs with TEE/ZK so Base wants 1 day just so they have time to monitor and react to such bugs.
On the other hand, Base also has a Solana canonical bridge. Here a set of validators (including Base and Chainlink) wait for Base's state update to be posted to the L1, then wait for the state update to achieve L1 finality, which takes about 15 min. Then, the validators simply sign and attest to the state update. On the Solana side, the smart contract verifies the validator signatures and then release the funds.
In short, with the canonical bridges, Base -> Ethereum takes 1 day to settle, while Base -> Solana takes about 15 minutes. This is quite ironic considering Base is an Ethereum L2, so you would expect Base to interop faster with Ethereum than Solana, but it's just the opposite.
Of course, the speed for the Solana path is not achieved without any cost. You are relying on the additional trust assumption that the validators behave honestly -- if the validators are hacked, they can attest to any invalid Base state. In other words, the Base->Solana bridge is a "multisig bridge." On the other hand, the Base -> Ethereum bridge is trust-minimized -- hacking the nodes is not enough; you must also break the TEE and the ZK proving program, which is much harder.
But still, at Ethlabs we believe that Ethereum chains -- including L1 and L2s -- must form a seamless "interop zone" where tokens flow securely AND quickly between networks, and 1 day is just not good enough. That's why we are building technologies that help tokens move way faster between L1 and L2s, without sacrificing security. Stay tuned!
J-pEgG🇬🇧ZKEVM🥷03/01/09Protocols are not app's👊 retweeted
This direction is what I'd dub "Private trains, public rails". The goal is the best network for assets to be issued, held and used, with value accruing to the network as a whole and ETH.
Ethereum is a confirmation engine. CCIP v1 already waited for Ethereum finality, but Chainlink nodes still had to report the result themselves. That is the pragmatic choice when proving Ethereum's consensus on another chain isn't practical. With zk-powered light clients, the bridge can verify Ethereum's confirmation directly instead of relying on someone's report of it.
Plugging these proofs into the bridge is enabled by the modular architecture of CCIP 2.0, also present in LayerZero. It unbundles the transport layer of the bridge from its application services, such as token issuance, standards and business logic, compliance or other features.
But it's only an attractive proposition if the Ethereum settlement layer is fast, cheap and secure, otherwise the temptation is great to use an alternative with different trade-offs. FCR is another step in this direction.
Zooming out, it starts to show what is imo the right balance between private and public build-out of Ethereum.
- Canonical bridges tying L2s to L1 remain the least featureful and most neutral, but the most secure way to move assets between L1 and L2s, especially ETH and other L1-native assets. They are privately built, but function more like public utilities: available, unattended, and secure, built on L1's settlement (finality or fast confirmation) and more and more on broadly available zk proofs.
- The utility exposed by these bridges becomes the root of trust for private operators offering differentiated services to asset issuers and other customers. Chainlink and others are well placed to serve their customers' business needs, relying on Ethereum's confirmation engine for the rails on which assets run.
Ethlabs is helping to build the public rails: fast, cheap and secure settlement across L1 and L2s, with value accruing to Ethereum and ETH.
It’s very cool to see more zk tech hardening bridges! zk does two things here: it proves that a fact of a world in a remote chain is part of that chain's state, and that this fact of the world is the one everybody agrees happened in the first place, ruling out double spends.
This is already a major improvement, and I'd like to talk about where we could go even further..
Today, and in this @chainlink CCIP 2.0 module, the @SuccinctLabs light client proves that Ethereum finality was achieved (“everybody agrees this version of the world happened”), by proving the report from a small validator subcommittee known as the “sync committee”. Finality is the ultimate agreement. However Ethereum finality still takes about 15 minutes!
Meanwhile, FCR is a snapshot of that agreement with very strong guarantees, at low latency. Now what if you could put together the power of zk and the power of FCR together to reduce by 30x that latency? Turns out, this may be possible! We mentioned it briefly in our weekly update two weeks ago teapawt.net/ox_shaman/status/21016…
Can we go even further? There are two complementary boosts:
1. Extending state inclusion proofs to full validity proofs. You're then not just proving that the state reported by validators includes data about your bridge transfer. You're proving validity of the entire remote chain's execution! No chance for validators to fool you.
2. Reducing time-to-finality. FCR is great, and will be a great bridge (pun intended) until we get really low latency finality, in the order of tens of seconds. Eventually, we'll get that, as well as a full proof that this finality was achieved, not relying on a subcommittee's report.
You put all this together, and you start to see the future of seamless and secure interoperability across Ethereum. Let's go build it!




























