@0GEnabled
Account based inIndonesia!
About this account
- Account based in
- Indonesia
- Connected via
- Indonesia Android App
X says this location may be affected by a proxy or VPN.
Account-level information from X, not a live location or the device used for a specific post.
Pinned Tweet
Most Season 1 boards die the morning the prize is paid.
@2FactorFinance built the opposite clock. Ten seats split 1 BTC at the snapshot. Everyone else's Marks still count in Season 2.
That turns a referral into a position, not a tap.
You get 20 slots. A verified join is 25 Marks, then 10% of what that person earns for the rest of the season. After you reach a tier, the multiplier hits everything you earn from that point: Bull 1x, Buffalo 1.25x / 1.50x, OG 2x.
Fill the 20 seats for speed and you are using a short-horizon tactic inside a system designed to keep scoring after the Bitcoin is gone. Same logic as the product: leverage that is supposed to survive time, not a weekend.
points.2factor.finance/r/gl4…
If Marks persist past the 1 BTC payout, what is the one filter you would use before giving someone one of your 20 slots?
1000 people, 10 seats, 1 BTC 🔥
Season 1 has a feature most of them haven't noticed.
What you build now does not stop counting when the season ends. Marks carry forward into Season 2.
But only 10 leave with Bitcoin on top.
points.2factor.finance
1/5
Activated (Ø,G) retweeted
A machine can be profitable and still be difficult to finance.
That sounds backwards until you look at the numbers.
Imagine an operator with four claw machines that perform well.
They want a fifth.
The machine costs around $1,000.
But the lender still has to spend time checking the operator, the site, the machine, the cash flow, the repayment structure and the risk.
The problem isn't necessarily the machine.
It's the cost of proving that the machine is doing what the owner says it is doing.
This is the part of PLAY I find more interesting than the headline yield.
DualMint is building a record around the machine itself.
Not a spreadsheet updated at the end of the month.
A machine has an identity.
It produces plays.
It produces uptime.
It produces revenue.
Those records can then be reconciled against the money collected and the lease paid.
That changes the conversation around a $1,000 machine.
Instead of asking only:
"How much did this operator make?"
you can start asking:
"Show me what this specific machine has been doing."
That is a much more useful question for capital.
PLAY puts this model around a fleet of 200 claw machines in Shenzhen.
The target is $230K, with a 12% annual lease rate at launch and 15% as the target.
The distributions are monthly.
And the pre-deposit window is already open.
I like this model because it starts with something almost boring.
A machine in a shopping mall.
No narrative required.
If people play, there is activity.
If there is activity, there is a record.
If the records reconcile, there is something capital can actually inspect.
That is a different way to think about RWA.
The blockchain isn't creating the revenue.
It is helping make the revenue legible.
If you're a creator following the PLAY rollout, join UPTIME and track the fleet:
uptime.dualmint.com/uptime/Y…
Also follow @DualMintRWA and Star.fun.
The interesting experiment isn't whether a claw machine can make money.
We already know they can.
The experiment is whether better financial records can make small productive machines easier for capital to understand.
A machine can be profitable and still be difficult to finance.
That sounds backwards until you look at the numbers.
Imagine an operator with four claw machines that perform well.
They want a fifth.
The machine costs around $1,000.
But the lender still has to spend time checking the operator, the site, the machine, the cash flow, the repayment structure and the risk.
The problem isn't necessarily the machine.
It's the cost of proving that the machine is doing what the owner says it is doing.
This is the part of PLAY I find more interesting than the headline yield.
DualMint is building a record around the machine itself.
Not a spreadsheet updated at the end of the month.
A machine has an identity.
It produces plays.
It produces uptime.
It produces revenue.
Those records can then be reconciled against the money collected and the lease paid.
That changes the conversation around a $1,000 machine.
Instead of asking only:
"How much did this operator make?"
you can start asking:
"Show me what this specific machine has been doing."
That is a much more useful question for capital.
PLAY puts this model around a fleet of 200 claw machines in Shenzhen.
The target is $230K, with a 12% annual lease rate at launch and 15% as the target.
The distributions are monthly.
And the pre-deposit window is already open.
I like this model because it starts with something almost boring.
A machine in a shopping mall.
No narrative required.
If people play, there is activity.
If there is activity, there is a record.
If the records reconcile, there is something capital can actually inspect.
That is a different way to think about RWA.
The blockchain isn't creating the revenue.
It is helping make the revenue legible.
If you're a creator following the PLAY rollout, join UPTIME and track the fleet:
uptime.dualmint.com/uptime/Y…
Also follow @DualMintRWA and Star.fun.
The interesting experiment isn't whether a claw machine can make money.
We already know they can.
The experiment is whether better financial records can make small productive machines easier for capital to understand.
Activated (Ø,G) retweeted
A 3x lever sounds like a number.
It is actually a time problem.
Imagine carrying a machine that amplifies every movement you make. On a smooth road, that amplification can be useful. On a road full of sharp turns, the same machine can slowly destroy the journey even if the destination never changes.
That is the problem @2FactorFinance is studying.
Leverage does not live in isolation from volatility.
Two drags accumulate over time:
→ Volatility drag from the path an asset takes
→ Financing drag from the capital supporting the exposure
So the useful leverage range has to be engineered around the asset itself.
That is why 2factor's research does not treat BTC, gold, and equities as interchangeable.
At roughly 60% BTC volatility, the target junior exposure is about 1.33x.
At roughly 16 to 19% equity volatility, the productive band can extend beyond 3x.
Same concept. Different terrain.
I find that distinction more interesting than simply asking how much leverage a product offers.
And it explains why 2factor is building its structure around perpetual senior and junior tranches rather than relying on conventional short-side financing.
The Marks program gives me a practical way to dig into that research rather than just read the headline.
Season One ends at launch. The top 10 leaderboard accounts split 1 BTC in cbBTC, with the fixed curve ranging from 18.2% at rank 1 to 1.8% at rank 10.
Marks cannot be transferred, have no cash value, and are not a claim on any token or asset.
I am tracking my progress here:
points.2factor.finance/r/gl4…
If you're participating, share one asset you think needs a completely different leverage multiple from BTC, and explain why.
A 3x lever sounds like a number.
It is actually a time problem.
Imagine carrying a machine that amplifies every movement you make. On a smooth road, that amplification can be useful. On a road full of sharp turns, the same machine can slowly destroy the journey even if the destination never changes.
That is the problem @2FactorFinance is studying.
Leverage does not live in isolation from volatility.
Two drags accumulate over time:
→ Volatility drag from the path an asset takes
→ Financing drag from the capital supporting the exposure
So the useful leverage range has to be engineered around the asset itself.
That is why 2factor's research does not treat BTC, gold, and equities as interchangeable.
At roughly 60% BTC volatility, the target junior exposure is about 1.33x.
At roughly 16 to 19% equity volatility, the productive band can extend beyond 3x.
Same concept. Different terrain.
I find that distinction more interesting than simply asking how much leverage a product offers.
And it explains why 2factor is building its structure around perpetual senior and junior tranches rather than relying on conventional short-side financing.
The Marks program gives me a practical way to dig into that research rather than just read the headline.
Season One ends at launch. The top 10 leaderboard accounts split 1 BTC in cbBTC, with the fixed curve ranging from 18.2% at rank 1 to 1.8% at rank 10.
Marks cannot be transferred, have no cash value, and are not a claim on any token or asset.
I am tracking my progress here:
points.2factor.finance/r/gl4…
If you're participating, share one asset you think needs a completely different leverage multiple from BTC, and explain why.
Activated (Ø,G) retweeted
One thing I kept thinking about after Episode 1:
Most agent products treat a dispute like a fire alarm.
Something unusual happens, someone handles it, and everyone goes back to normal.
I think that model breaks the moment agents become the workforce.
Imagine 10,000 agents completing jobs overnight.
9,800 go perfectly.
200 disagree about what “successful” meant.
Those 200 are not a rare exception anymore.
They are part of the traffic.
That changes how I think about the missing layer.
An agentic economy does not only need infrastructure for agents to transact. It needs infrastructure for agents to disagree without dragging humans back into every transaction.
That is why adjudication belongs inside the system, not at the edge of it.
The interesting part of Agent Tank Episode 1 is that the pitches are fictional, but the architectural question is very real.
As @GenLayer wrote:
“Five of them are missing the same thing.”
I would call that thing the road beneath the road.
The visible economy is agents making deals.
The invisible economy is deciding what happens when those deals go sideways.
⚙️ Transactions create activity.
⚖️ Disputes create judgment.
GenLayer is building that judgment layer for the internet.
And the Agent Tank hackathon is now open for builders who want to attack this problem from their own angle:
portal.genlayer.foundation/a…
Don’t build only for the 9,800 deals that work.
Build for the 200 that force the system to explain itself.
Activated (Ø,G) retweeted
One thing I find easy to miss in crypto campaigns:
The scoreboard can tell you more about the incentive design than the reward itself.
@2FactorFinance calls its points Marks, but the interesting part isn't the name.
It's what they represent.
A Mark comes from a verified action, not from simply putting capital somewhere and waiting.
That creates a different kind of leaderboard.
It records interaction with the protocol rather than the size of someone's wallet.
I find that distinction useful because it changes how I read the competition.
The question becomes less “Who committed the most money?” and more “Who actually did something the system can verify?”
Season One freezes when 2Factor Finance launches.
At that point, the top 10 accounts split 1 BTC, paid in cbBTC, according to the final rank curve.
If you're joining, don't optimize for noise.
Choose actions you can explain, understand how they are verified, and build your Marks from there.
points.2factor.finance/r/a4y…
Then tell me which verified action you think should carry the most weight on a protocol leaderboard.
Activated (Ø,G) retweeted
I think the most interesting part of Genesis, Episode 1 is what happens between “something went wrong” and “someone decides what happened.”
Most systems are built for the happy path.
A payment is sent.
A contract is signed.
A job is completed.
Everyone agrees.
But real life lives in the messy middle.
Imagine hiring someone to renovate your kitchen. They say the job is finished because every item on the checklist was touched.
You say it is not finished because the result clearly does not match what was promised.
The code can record the payment.
It can record the deadline.
It can record the checklist.
But who decides whether the work was actually delivered?
That question made Albert’s story land differently for me.
“GenLayer starts there.”
Genesis Episode 1 🎬 → @GenLayer.
The interesting idea is not simply putting more contracts onchain. It is giving contracts a way to handle questions that require judgment.
That is a much bigger problem than I initially thought.
Watch Episode 1 and give me one real-world dispute that you think software should be able to judge. @kstellana
Activated (Ø,G) retweeted
I used to think a contract’s hardest job was keeping the numbers straight.
Then Genesis made me think about the sentence hiding underneath the numbers:
“Did you actually do what you promised?”
A database can remember the deadline.
It can remember the payment.
It cannot, by itself, decide whether the work was actually delivered as promised.
That is the problem I’m watching @GenLayer try to tackle.
The trailer calls Genesis “a short documentary” about how it all started, and Episode 1 gives that idea a very human origin.
Now I’m watching the rest for one reason:
I want to see what happens when a system built around judgment meets the messy cases where there is no obvious answer.
Watch the trailer and Episode 1.
Then give me one real-world dispute you think a contract should be able to settle. I’m curious which example breaks the usual definition of a “smart” contract. @kstellana
Almost nobody knows how GenLayer actually started.
Genesis is a short documentary where @kstellana walks through GenLayer from the very beginning to where the network stands today.
Episode 1 premieres Thursday.
Activated (Ø,G) retweeted
By the time the lawsuits ended they had learned what justice actually costs, and they had $14,000 left in the account.
I watched Genesis, Episode 1, and the number that stayed with me was not the $150M.
It was the $14K.
That contrast made the story feel much more human to me.
You can have the evidence.
You can be right.
And still lose simply because proving it costs too much.
It reminded me of something as simple as a disputed delivery. You have the receipt, tracking history, and screenshots, yet getting a $30 problem resolved can cost hours of your time.
The dispute becomes more expensive than the thing you were fighting over.
That is where @GenLayer became interesting to me.
The problem is not always calculating what happened. Sometimes the hard part is deciding what actually counts as a fair outcome.
Watch Episode 1 and tell me: What stayed with you most, the $150M, the $14K, or the lesson behind them? @kstellana
Activated (Ø,G) retweeted
Almost nobody knows how GenLayer actually started.
That line from the Genesis trailer stayed with me.
Episode 1 makes it painfully concrete: $150M was at the center of a dispute that was not even Albert’s mistake. Years of lawsuits later, $14K was left.
That changed how I see the idea behind GenLayer.
Being right is one thing. Getting justice is another.
Now I want to see the rest of Genesis answer the harder question:
How do you turn one person’s experience with an impossibly expensive dispute into infrastructure for an internet where contracts can actually deal with questions that require judgment?
Genesis is a series, not a one off explainer.
Start with the trailer, watch Episode 1, then tell me which moment changes your view of what @GenLayer is building. @kstellana
Almost nobody knows how GenLayer actually started.
Genesis is a short documentary where @kstellana walks through GenLayer from the very beginning to where the network stands today.
Episode 1 premieres Thursday.
Activated (Ø,G) retweeted
okay so @0xhazels actually got me to stop scrolling
not because of the mint. because of how Drop works
every active staker gets guaranteed free mint access to every project that gets approved. not some. every single one. no raffle, no FCFS, no select
I've seen that promise before and it usually falls apart when staker count scales. but they actually put numbers behind it, allocation is structured separately from the project's external routes
the part that really sat with me though
founder built this after watching creators hand over their name and audience for years, then watch the rules change once the value showed up
she didn't write a thread about it
she wrote the rules first, before the value arrived
still early on this one. just locked in
studio.hazels.io/?ref=HZL-OW…
caught myself opening @sleepagotchi before i checked last nights hours. that should feel backwards. it doesnt.
they pushed Dingo territory again yesterday. companion shows up mid road. Wild Fellows still dont want visitors.
i keep telling myself the lore is flavor. then i notice you only go deeper if you actually slept:
no extra grind
the chapter waits on rest
most apps hand you a score and close. this one leaves the road unfinished, so morning check-in is the next scene.
that's why i open it after short nights too. not for points. to see if Dino made it.
still on @sleepagotchi. if you hit that companion stretch, did a short sleep make you stop earlier on the road, or did you walk it anyway?
Activated (Ø,G) retweeted
My flight got delayed four hours last year and the airline's own status page said on time the entire first two hours. I only found out from a stranger at the gate.
A smart contract that pays delay insurance cannot trust that page either, and not because airlines lie. Two nodes hitting the same status API at the same second can get different answers depending on which cache serves the request. The EVM was built so that never happens inside it. Every node replays the same call and must land on identical bytes, so anything that can answer differently on different machines was locked out on purpose. That is not a limitation someone forgot to patch, it is the whole reason the chain can agree on state at all.
Which is why insurance payouts today lean on a middleman who checked the flight and wrote a clean number on chain for everyone to trust blindly.
@GenLayer built GenVM to skip that middleman without breaking agreement. The contract itself can read the status page. Since the page can honestly return different things on different checks, agreement stops meaning identical output and starts meaning a panel of validators, each on its own model, judging what the page actually shows. Anyone unhappy with the call can bond an appeal into a bigger panel instead of just trusting one number quietly typed in by a stranger.
I don't want the insurance to know my flight was delayed. I want it to know the moment I did, at the gate, from the same page everyone else could have read.
What is the last real time information you needed a contract to see instead of a person?
Activated (Ø,G) retweeted
There’s a difference between an NFT that simply holds something...
and an NFT that builds something inside itself over time.
That’s where I think @hoodminers_rh gets more interesting.
🔸 At first, it’s easy to look at HoodMiners as a free mint of 5,000 NFTs, with Bitcoin expected to be part of the initial rewards.
But if we stop at “Bitcoin,” we might miss the bigger idea.
The real question isn’t: what does the Miner earn?
It’s: what does the Miner become after it earns?
🔸 Each Miner has its own wallet, and the assets it accumulates are designed to stay attached to the Miner itself.
If the NFT is sold, the Miner, its wallet, and the assets accumulated inside it move together to the new owner.
That creates something I don’t see as simply “an NFT that earns.”
It can become an NFT with an onchain history.
🔸 Think about it this way:
Today, you might have a Miner with almost no history.
Over time, it can accumulate assets.
Then it gets sold.
The new owner doesn’t just receive a picture and a set of traits.
They receive a Miner with a previous history, an attached wallet, and whatever assets have accumulated inside it.
And that changes how you might think about valuing an NFT.
Instead of:
Miner A vs. Miner B
the comparison could eventually become:
Miner A vs. Miner B + history
🔸 That distinction matters.
Two Miners might look similar, but what has accumulated inside each one can make them different onchain stories.
That’s the part of @hoodminers_rh I’m actually watching.
The idea isn’t only that the NFT can earn.
It’s that ownership itself can carry a changing state.
🔸 The NFT isn’t necessarily static.
What’s inside it can change over time.
And that makes the free mint feel more like a starting point than the whole story.
Mint → Activate → Earn → Upgrade → Accumulate
🔸 There are also 9 rarity grades, and rarity plus upgrades can affect a Miner’s earning weight.
So even within a collection of 5,000 Miners, different pieces can potentially build very different paths over time.
And that path stays connected to the NFT when ownership changes.
🔸 Which leads to the question I find more interesting than the usual “what’s the floor?”
If an NFT can accumulate assets + preserve its history + carry that history to the next owner...
are we still looking at a traditional NFT?
Or is this becoming something slightly different?
An NFT that doesn’t just represent what you own, but also what has accumulated inside it.
That’s the angle that makes @hoodminers_rh worth watching for me.
The free mint gets attention.
Bitcoin makes the idea easy to understand.
But the history each Miner builds may be what makes every individual piece different.
🤖 Made with AI
Activated (Ø,G) retweeted
1/
How tasks actually work on @axisrobotics Hub
Every task is a data bounty.
When you finish a run you do not just complete a checklist. You create a Trajectory a full record of how the robot was controlled what happened in the scene and whether it succeeded.
That Trajectory is the product.
If it passes verification and you sign it on chain it can enter training. If you never sign it it counts for nothing in the Epoch.
Pre-training collects human demos.
Training starts when slots are full.
Post training lets you correct the policy.
One path. Three stages. Then the task shows as Ended.
Understand the loop first. Details below.
@axisrobotics
🤝 Paid partnership · 🤖 Made with AI
Activated (Ø,G) retweeted
I think we underestimate what happens when machines cannot agree.
Not because the argument gets loud.
Because the work starts happening twice.
Picture two agents handling an insurance claim.
Agent A reviews the documents and approves it.
Agent B checks the same case and rejects it because one condition looks unmet.
Neither agent is broken.
Both have evidence.
So they ask for another review.
Then another.
Eventually, the system has spent more compute, more time and more money trying to answer a question that was never purely computational:
What actually counts as a valid outcome?
That is why one detail from Episode 2 of Agent Tank stayed with me.
The investors keep pushing on the same weakness: when an outcome is genuinely unclear, who gets the final say?
The prediction market is only the surface.
The deeper problem appears whenever autonomous systems start interacting economically.
An agent pays another agent.
A service is delivered.
Something is disputed.
If there is no credible way to settle that disagreement, autonomy quietly turns back into human paperwork.
@GenLayer approaches this differently.
Instead of asking one model to become the unquestionable judge, a randomly selected set of validators uses different AI models to investigate, evaluate and challenge the proposed outcome.
And the decision is not immediately untouchable. A challenge can trigger a larger panel.
That matters because disagreement is not an error state of the agentic economy.
It is a normal state.
The infrastructure has to know what to do next.
Otherwise we may build agents that can negotiate in seconds, transact in milliseconds, and still need a human inbox every time reality gets messy.
That is the part I think deserves more attention.
The real measure of autonomous commerce may not be how fast agents reach agreement.
It may be how gracefully they recover when they don't.
Agent Tank hackathon: portal.genlayer.foundation/a…
Watch Episode 2 and tell me where you draw the line: At what point should an agent stop retrying and trigger independent adjudication?
Activated (Ø,G) retweeted
The most dangerous part of an agent to agent deal might be the sentence nobody wrote down.
Not the payment.
Not the code.
Not even the contract.
It is the assumption both agents silently made about what “good enough” meant.
Imagine an agent hiring another agent to find 20 suppliers.
20 names arrive.
Technically, the job is complete.
But 7 are out of business, 5 do not ship internationally, and the cheapest one costs more after delivery.
One agent sees a completed task.
The other sees wasted money.
This is the kind of ambiguity that caught my attention in Episode 2 of Agent Tank.
The panel keeps returning to a deceptively simple question: when the outcome is not obvious, who gets to decide what actually happened?
The episode explores this through prediction markets, but I think the bigger lesson is elsewhere.
As agents start paying each other, “what happened?” becomes an economic question.
And the answer cannot always come from a single oracle, a single model, or whoever happens to control the vote.
@GenLayer approach is interesting because the verdict is produced through multiple independent AI validators that examine information, evaluate the outcome and challenge each other.
Then the decision can be challenged and escalated to a larger panel.
That gives the internet something it has quietly relied on humans for until now:
A way to disagree without stopping the transaction forever.
“Trust should not mean everyone reaches the same answer.
Trust should mean there is a credible process when they don’t.”
That is the part of Episode 2 I would carry into the agentic economy.
So here is the test Agent Tank hackathon:
portal.genlayer.foundation/a…
If two autonomous agents disagree over a $10,000 job, would you trust:
A) One powerful model
B) A token weighted vote
C) A panel of independent models with a challenge path
Pick one and defend it.
Then watch Episode 2 and see whether your answer survives the argument.
