milo retweeted
Do not bother. You are too late. A barber told me that in 2019, clippers still running, and I obeyed him for two years.
He had owned bitcoin once and sold it. That made him sound like a witness, not a guesser.
So I watched from the side. I called it discipline. It was obedience.
In November 2021 I bought 4,000 dollars of everything, at the top, in one evening.
Not because I believed in the coins. Because I wanted to stop being late.
By March it was 300. I checked it in his chair under the cape, a tab away from @RallyOnChain.
Last month I sat in his chair again. I told him he was right about everything.
He asked which coin. Then he asked about the match last night. He had forgotten the sentence.
The danger was not that he was wrong. It was that he was casual.
He was filling silence. I was taking notes.
If you are still trading against a sentence that was not yours, tell me what it cost you and what you do differently now. I will read every answer.
milo retweeted
A film project’s pitch can look great. I’d still check its team vesting before joining through @IPXExchange.
IPX gives creators a tokenomics setup that goes beyond choosing a token price.
The process starts after project approval:
1. Set the total token supply and default vesting terms.
2. Add each team member’s wallet, role, token allocation, and vesting schedule.
3. Check that allocations do not exceed the total supply and that the cliff does not exceed the vesting duration.
4. Save the configuration and request token deployment.
There is a useful detail for film projects: IPX’s Film / Creative template recommends 90-day vesting with no cliff. It is a suggested template, not a mandatory rule for every project.
For example, on a hypothetical 1 billion token supply, a 100 million team allocation represents 10%. I would want to know who receives that allocation and when those tokens unlock, not just the headline supply.
The timing matters too. Creators can edit the configuration before deployment. Once deployed, the vesting schedule is locked on-chain.
My checklist as a potential member is simple: team allocation, wallet recipients, unlock duration, cliff, and total supply.
A project’s story tells me why it deserves attention. Its tokenomics help me judge whether the incentives are structured clearly.
milo retweeted
Dear whoever bought my 3.2 ETH at $1,082.
June 2022, 4:06 a.m., kitchen light off so the screen wouldn't wake my wife. A thread said the lenders were insolvent. I believed a stranger over my own spreadsheet and hit market sell. $3,462.
You got filled instantly. Someone thinks this goes up, I thought. I needed you to be an idiot so I could feel like the smart one.
I went back to bed and told her I'd trimmed exposure. That wasn't a trim. It was all of it, and half was hers.
The sell wasn't the mistake. The word was. Trimmed. De-risked. Clean words kept me from saying the true sentence out loud: I got scared.
I've wanted to write to you since that night. A market order has no return address. You never find out who took the other side, so you never have to explain yourself to them.
I'm leaving this here for @RallyOnChain's letter campaign because you don't have an inbox.
You had a word too. Tell me yours. I want to know if it sounded as reasonable as mine.
If you're standing in that kitchen: say the plain sentence to whoever shares the money before you press sell. If you can't say it plainly, you're not managing risk. You're hiding.
🤝 Paid partnership
Do not bother. You are too late. A barber told me that in 2019, clippers still running, and I obeyed him for two years.
He had owned bitcoin once and sold it. That made him sound like a witness, not a guesser.
So I watched from the side. I called it discipline. It was obedience.
In November 2021 I bought 4,000 dollars of everything, at the top, in one evening.
Not because I believed in the coins. Because I wanted to stop being late.
By March it was 300. I checked it in his chair under the cape, a tab away from @RallyOnChain.
Last month I sat in his chair again. I told him he was right about everything.
He asked which coin. Then he asked about the match last night. He had forgotten the sentence.
The danger was not that he was wrong. It was that he was casual.
He was filling silence. I was taking notes.
If you are still trading against a sentence that was not yours, tell me what it cost you and what you do differently now. I will read every answer.
milo retweeted
The EVM is a courtroom, not a calculator. It rules on agreements, but it grinds numbers slowly.
Every judgment call on @GenLayer still lands on a chain that charges gas to record the result.
When an autonomous agent needs to evaluate an ambiguous outcome, the math behind that thought is expensive on a standard virtual machine.
The GenLayer Chain now runs on the @arbitrum stack. That stack ships Stylus beside the EVM.
Stylus runs contracts written in Rust. They compile to WebAssembly before execution.
The rules stay in Solidity.
The heavy math can move to Rust.
The chain hosts both engines.
Training the model is not the hard part of onchain AI. Paying the gas bill for every thought in production is.
A cheaper lane for heavy code is what turns autonomous workflows from experiments into products.
Same courtroom, new calculator.
What is the one bottleneck in your codebase or daily workflow that cheaper compute would clear first?
milo retweeted
Imagine losing access to millions because an agentic contract had no courtroom to turn to.
Traditional legal systems move at the speed of years while autonomous scripts execute at the speed of light.
That exact tension is what stood out to me while listening to @kstellana talk with @HenriArslanian.
When AI agents negotiate deals globally, human judges cannot keep up.
The bottleneck is not writing code anymore.
Execution has become virtually free across every major network.
The expensive part is judgment.
Deciding who pays when a subjective clause fails is where @GenLayer steps in with decentralized adjudication.
If a protocol cannot resolve ambiguity on chain, the entire agentic economy stalls.
Builders cannot scale trust by relying on outdated litigation.
So the question is simple.
Would you trust an automated dispute layer with your treasury today?
One line is enough.
I read every reply.
milo retweeted
You cannot manufacture ten years of production scars. You can only stand on someone else's.
Nitro has years of production use behind it, mainnet failures other people already survived. That is what @GenLayer is buying with this move.
Watch any team you trust: the least interesting part of their stack is usually the part somebody else already broke and fixed.
Geth is the client most Ethereum tooling already speaks to, so the standard kit fits the chain as it is, without adapters.
Stylus is a second virtual machine beside the EVM, running Rust contracts compiled to WebAssembly. The consensus contracts stay in Solidity today. It is an option they are not exercising yet.
The part that moved is the GenLayer Chain: the contracts recording transaction order, validator committees, votes, appeals, staking and final outcomes. They left the ZK Stack for Arbitrum Nitro, operated with Conduit.
What cannot be borrowed is the judgment. Optimistic Democracy and GenVM work exactly as before. Conduit runs the infrastructure; the team works on the protocol. Matter Labs stays on Prividium, the private deployments for institutions.
Edgars Nemše walked through the reasoning on 10 November, and it holds: @GenLayer did not change what it decides, only where the record of this @arbitrum move lives.
Name the last thing you chose to rent instead of build, and how many years of someone else's failures came with it.
milo retweeted
AI agents can close deals at machine speed. When they disagree there is still nowhere for them to go.
@kstellana explains the gap on @HenriArslanian’s show. Traditional courts assume a human is on the other side. Agents do not fit that system. You cannot serve them or wait months for a judgment when the transaction already settled.
@GenLayer is building the adjudication layer that can handle it. Independent validators running different models reach consensus on the disputed facts and enforce the outcome on-chain.
Would you trust a panel of models to resolve a commercial dispute your agent entered or do you still want a human in the loop for anything material?
still here 💚
@TxFlow_L1
2
31
1,684
milo retweeted
A second VM isn't a reason to rewrite working contracts. It's another option when the right problem comes along.
One detail worth paying attention to in @GenLayer's move to @arbitrum is Stylus.
GenLayer's consensus contracts are written in Solidity today. Nitro also has Stylus, which lets contracts written in Rust run alongside the EVM through WebAssembly.
That gives GenLayer room to keep its existing coordination logic in Solidity while potentially moving compute heavy parts of the protocol to Stylus later.
I think the important part is being selective. Moving code to a different VM only makes sense when there's a real reason to do it.
I'd rather see a protocol use a new execution environment where it solves a specific problem than rewrite things just because the option exists.
Which part of a protocol would you consider moving out of the EVM first?
Why @TxFlow_L1 Stands Out vs Hyperliquid, Arcus, Aster & Lighter
1/ While most new perp DEXs compete on fees or leverage, @TxFlow_L1 is building something broader: a purpose-built Layer 1 for all finance.
Their vision: “the blockchain where all finance happens.”
2/ What is TxFlow?
Fully on-chain CLOB (central limit order book) for spot, perpetuals, and prediction markets
Tx-BFT consensus (DAG-based BFT) → parallel execution, no single-leader bottleneck, non-reversible finality
TxCore (trading engine) + TxEVM (native EVM running in the same consensus stream)
TIP Liquidity Standards → shared liquidity across different “Channels”
Every order, cancel, fill, and liquidation is a public, replayable transaction. Nothing relies on trusting a single operator.
3/ Vs Hyperliquid (the clear volume & OI leader)
Both run fully on-chain order books on their own L1s with nearly identical base fees (0.015% maker / 0.045% taker).
TxFlow differentiates with:
TIP shared liquidity layer so new products don’t start from zero
TxEVM integrated into the same consensus (asset issuance + smart contracts live alongside the order book)
Hyperliquid is far more mature and liquid. TxFlow is architected for a multi-product financial ecosystem from day one.
4/ Vs Arcus (dYdX team on Robinhood Chain)
Arcus uses a hybrid model: off-chain matching + on-chain settlement/verification.
TxFlow is fully on-chain end-to-end — matching, margin, and liquidations all happen inside a deterministic state machine.
Higher transparency and atomic consistency. Arcus is stronger on RWA/stock perps and multi-asset collateral, but carries more trust assumptions.
5/ Vs Aster
Aster has its own L1, aggressive leverage (up to 1001x in special mode), and often 0% maker fees.
TxFlow prioritizes full verifiability and standardized shared liquidity over extreme leverage or privacy features.
Aster leans privacy-focused with multi-chain deposits; TxFlow keeps everything in one replayable consensus.
6/ Vs Lighter
Lighter’s big edge is true zero fees for standard accounts + ZK proofs anchored to Ethereum.
Trade-off: higher latency (~300ms taker on the free tier).
TxFlow targets high throughput (claimed 250k TPS) with one-block finality via its DAG-based consensus, plus native multi-product composability and TxEVM.
Lighter wins on pure cryptographic verifiability; TxFlow wins on unified liquidity and broader financial surface.
7/ Core differentiators
TIP Shared Liquidity → new Channels inherit existing liquidity
TxEVM in the same consensus → assets can be minted on the same chain as the order book that trades them
No VC token allocation → community-owned from the start
Extreme transparency: anyone can replay the chain and verify every fill
8/ Current status (as of Nov 2026):
Cumulative perp volume > $12B
15,000+ traders, Points Season 1 live
Fiat on-ramps, mobile app, prediction markets, stock/forex/pre-market perps
Still invite-only via Discord
USDC bridge (OpenZeppelin audited), moving toward native issuance with TxEVM
Volume and open interest are still significantly smaller than Hyperliquid, Lighter, and Aster this is early stage.
9/ Quick comparison:
AspectTxFlowHyperliquidArcusAsterLighter
MatchingFully on-chainFully on-chainOff-chainOwn L1ZK on-chain
Shared LiquidityTIP standardsHIP-3———
Native EVMYes (same consensus)HyperEVM——Ethereum L1
Base Perp Fees0.015% / 0.045%SameCompetitive0% maker0% (standard)
StageEarlyMatureGrowingGrowingStrong
TxFlow isn’t trying to out-volume Hyperliquid today. It’s building infrastructure designed for a more complete on-chain financial system.
If you care about full transparency + shared multi-product liquidity, it’s one of the more interesting early projects to watch.
DYOR, still very early.
@TxFlow_L1 #TxFlow #PerpDEX #OnchainFinance
milo retweeted
Switching phones taught me what a clean migration feels like: nothing missing.
My fear was never the hardware. It was my contact list, accumulated across years, even people I barely speak to now, plus WhatsApp chats and treasured pictures. After setup, each bit was there.
On 22 October, @GenLayer pulled off an identical achievement at tremendous scale, and @arbitrum is the new address.
The GenLayer Chain relocated, an EVM chain where consensus leaves its paper trail: the order of transactions, which committee decided every case, how it voted, any appeals, staking balances and the final result. It now lives on Nitro instead of the ZK Stack from Matter Labs, with Conduit handling operations.
Untouched by the trip:
• Optimistic Democracy, the method for reaching a verdict
• GenVM, the engine running Intelligent Contracts
• The validator job, with nodes pointing at a new destination
The real labour sat in three seams: the bridge, the link to Ethereum, and gas costs.
Reasons I find convincing:
• Arbitrum One ranks among the two biggest L2s by value secured, and over 30 purpose-built blockchains share its technology
• Underneath it all, Geth sits at the core, so familiar developer tooling integrates directly
• Its operator already runs more than 60 mainnets, liberating the team to concentrate on protocol development
Prividium stays in the picture: a private, permissioned deployment for institutions, say a bank consortium settling agreements internally. Bradbury testers received their following stage on that very day.
My read for an adjudication layer serving AI agents: judging should stay steady while infrastructure grows busier.
When did you last migrate something you could not afford to lose, what did you check first, and would you demand likewise for a court that settles disputes between autonomous software?
Arbitrum puts GenLayer near builders. A chain move won't tell an app what to do with a verdict.
@GenLayer is moving its public chain from ZK Stack to @arbitrum Nitro. The chain coordinates consensus, not AI execution.
Its contracts record committees, votes, appeals and outcomes. Optimistic Democracy and GenVM stay the same.
If two agents dispute whether a service met its terms, a payment alone can't say whether the work was done.
Validators using different AI models reach consensus on judgment calls. Intelligent Contracts execute in GenVM on their nodes.
Bradbury's next phase is live. ZKsync work continues on Prividium and enterprise, a separate path from the public network.
Arbitrum's ecosystem has teams GenLayer wants to reach. It's a chance to meet agent builders, not proof of adoption.
I'd start with the app: can it use the verdict to decide whether to pay? Validators still need to connect to the new stack.
Read Edgars Nemše's "Why GenLayer moved to Arbitrum". Which would you test first: validator migration or an app using a verdict, and why?
Dear the hardware wallet I never opened, you have lived in my drawer for fourteen months beside the passport copies. @RallyOnChain finally stamped this letter.
I bought you the night the exchange paused withdrawals, hands shaking, card declined once, then approved. The box went into the drawer.
I filed that under prudence.
A phishing mail said verify before midnight, and my hands shook for me. I opened the box at 1:40 and finished in 38 minutes, the number I cannot forgive.
The hard part was never the transfer. It was admitting the box was procrastination in costume. Security you postpone is not security, it is hope with a phrase.
So here is the line I owe you. If your backup sits sealed, do the transfer tonight, before the scare mail arrives, because fear is a bad scheduler.
Anyone else keeping a sealed box, what is the task and how many minutes does it really take? I read every reply and I answer the oldest box first.
milo retweeted
A verdict is just an opinion until someone can enforce it.
Picture two AI agents disputing a delivery that arrived "as described". Someone has to decide, and then someone has to make the loser pay.
Most pitches for machine justice stop at the first half.
My take: watch @kstellana give @HenriArslanian the second half. The held money releases itself, on whichever chain it sits. That is the part of @GenLayer's pitch I would bet on.
Tell me what made the last decision you disagreed with actually stick.
Full episode: youtu.be/uXpzHMc9mnw
My worst crypto mistake was two characters wide, shipped in daylight to a judge that reads everything. A trade loses money and this one lost a name.
Forty minutes before deadline, one draft left, public gratitude to the man who rebuilt my deck at 2 AM and never put a subject line on it. The rule said tag the person.
The bracket shipped.
The scoring read the bracket out loud in three places and the multiplier dragged every other line down by a sixth.
The judge was not human, and @RallyOnChain AI quoted my bracket back while an old friendship sat unnamed.
The hard part is not writing the gratitude. It is checking the field that carries a name, because everything else can bleed and still score and a name cannot.
Now every handle gets typed twice and nothing ships until both match. The attached card says type the name twice, in my own hand. Gratitude with a placeholder thanks nobody.
Confessors, my line one says type the name twice. What wrote your line one, and who almost went unnamed in your story? I read every reply and I answer the ones that name a person.
milo retweeted
I judge a protocol by what it refuses to operate itself. A good chef does not wash the dishes. The stove needs him.
One part of this move I keep pointing at is the operator. The @GenLayer chain now runs on @arbitrum Nitro, and Conduit operates it, not the protocol team.
Conduit already runs more than 60 mainnets. Running chains is its whole business, so the GenLayer team stays on the protocol, which did not change.
The hard part is not running a chain. It is keeping the people who judge from also being the people who reboot. Judgment and ops want different hands.
An adjudication layer sells judgment, not uptime. Uptime can be bought from a specialist. Judgment cannot. That is why the split matters more than the stack.
So this move is a hiring decision, not a speed decision. The chain got a specialist operator. Its protocol team kept its focus.
Builders, what would you hand a specialist operator tomorrow, the chain, the bridge, or the gas math? One line is enough. I read every reply.
milo retweeted
I stop trusting a reserve page that shows one column. My rule, count what they owe before what they hold.
The claim is fully backed.
After the 2022 blowups, exchanges published proof of reserves, a cryptographic list of every coin in their wallets, and readers called it solvency.
It is not solvency. Half a balance sheet.
The proof is a snapshot. Coins in named wallets at one timestamp, checked by a merkle tree. Real, and useful.
I read it for @RallyOnChain like a lease, assets on page one, debt on page two. These pages never print page two.
Page two is every deposit owed to users, every loan taken, every claim on them. Solvency is assets minus liabilities.
The picture is the two pages. Left full, right missing.
Once you see the missing column, the hard part is not reading the page. It is noticing the page that is not there.
Insolvency hides in the missing column.
So the question is not whether the coins exist. It is whether they are theirs or yours.
One question travels. Where is the half that shows what you owe?
Name a reserve page you trust and show me its liabilities line, or say it has none.
🤖 Made with AI



















