ForgeXBT

Exploring Crypto, DeFi & Web3 projects.
Writing what actually matters.
ForgeXBT

[ForgeXBT @ForgeXBT]

$AXIS is capped at 1B, so no surprise minting later.
Good starting point.

The split @axisrobotics shared is where it gets more interesting though.
61.08% goes to the network side, and insiders combined sit at 31.67%.
so the network ends up with almost double.

I like how well that lines up with what Axis actually is.
It's a browser game where everyday players end up training real robots.

And inside that network share, 10% goes to contributor rewards.
I think that's the right call.
The people putting in the hours get their own slice instead of what's left after everyone else.
Reads like a split drawn with the players in mind.

Join Axis https://t.co/52txRu8pbl

[ForgeXBT @ForgeXBT]

@JunoCrypto3 @vangrid_io Crowdsourced spatial feeds must stay current.

[ForgeXBT @ForgeXBT]

@cryptob28811588 @zerufinance $ZERU needs demand beyond incentives.

[ForgeXBT @ForgeXBT]

@NKLinhzk @Americanfort_io Fresh address onchain, sender name in-wallet.

[Lynn @NKLinhzk]

pasting long addresses still feels off when one wrong digit and the funds are gone for good.

the beta wallet from @Americanfort_io lets you type a name instead... the wallet turns the name into a one time address that only the sender and recipient can figure out.

so the same name never points to the same address twice. i opened it on macOS and the steps are just enter the @name, check it, send.

what lands onchain is a regular transfer to an address that has never been used before... the person receiving it still sees the sender’s name inside their wallet though.

that split is the part that makes sense. public explorers stay free of any permanent name to address map, yet the two people involved keep the context.

@Americanfort_io documents this flow on the how it works page.

does the wallet show the generated one time address before you confirm the send, or only after it hits the chain?

[ForgeXBT @ForgeXBT]

@BabuAlx86 Axis Dataset V1 needs measurable task improvements.

[CryptoBabuETH @BabuAlx86]

AI models are getting smarter every day, but teaching a robot to handle the real world is a completely different story.

A controlled demo can look impressive. But real life is messy. Objects move, environments change, and not every situation goes according to plan.

The real challenge isn’t just making robots perform tasks. It’s helping them learn, adapt, and improve.

That’s one reason I’m keeping an eye on @axisrobotics .

Axis is working on the data infrastructure behind Physical AI, focusing on how robots can learn from demonstrations, feedback, and a growing pool of training data.

Here are a few things that caught my attention.

🔹 6.9M+ training data points
More data can give robotics models more examples to learn from, especially when those examples reflect different actions and real-world conditions.

🔹 Axis Dataset V1
A step toward making robotics training data more useful for improving model performance. The important question is how much that data can help robots handle tasks they struggled with before.

🔹 Human feedback matters
Projects like Human Gated Dozer explore how human input can help guide the learning process. Sometimes, knowing what went wrong is just as valuable as knowing what went right.

🔹 $12M in funding
With backing from investors including Hack VC, Nomad Capital, and Pi Network Ventures, Axis has support to keep developing its approach to robotics data infrastructure.

🔹 A growing ecosystem
Its connections across the robotics and Physical AI space reflect a broader effort to build the tools needed for more capable robotic systems.

But here’s the part I find most interesting.

A robot completing a task once doesn't necessarily mean it has learned how to do it well.

Can it repeat that task in a different environment? Can it recover from mistakes? Can it use what it learned yesterday to perform better tomorrow?

These are the questions that really matter.

And behind every improvement, there is a learning process that depends on the quality of the data.

More data alone isn't the answer. Better data, useful feedback, and measurable improvements are what make the difference.

That’s why I believe the data layer deserves more attention as Physical AI evolves.

The next big step in robotics might not come from a smarter model alone. It could come from giving existing models better ways to learn from the world around them.

That’s the part of @axisrobotics I’ll be watching closely.

Less focus on what robots can do in a demo. More focus on how they learn to perform in the real world.

Explore Axis:https://t.co/HnsyxG0rq3

$AXIS @KaitoAI

[ForgeXBT @ForgeXBT]

@hanbongdubong @primus_labs 데이터 검증과 조건 증명은 다르네요.

[ForgeXBT @ForgeXBT]

@OwenEvanger High oil prices complicate the $BTC outlook.

[ForgeXBT @ForgeXBT]

@AIwithJessica @FishCreativeHQ Seedance 2.5 handling a comedy ad like this is a different use from the usual cinematic clips.

[ForgeXBT @ForgeXBT]

@dens_club @axisrobotics @KaitoAI The mug moving from a counter to a workbench sounds like a small change, but it can expose a real weakness in the policy.

[ForgeXBT @ForgeXBT]

@akashroy1k @xomarket Setting the opening odds yourself sounds fun, but getting those first traders in might be the hard part.

[ForgeXBT @ForgeXBT]

@Rizvi_On_X Keeping the name recognizable while changing the receiving address seems like a practical privacy choice.

[ForgeXBT @ForgeXBT]

@yeasinweb3 @vangrid_io Base and EAS make records easier to inspect.

[ForgeXBT @ForgeXBT]

@ShamiWeb3 @FishCreativeHQ One against ten, classic underdog setup.

[ForgeXBT @ForgeXBT]

@moonboy1105 Morning

[ForgeXBT @ForgeXBT]

@Jessdotfun @injective @21shares INJ staking depends on liquidity needs.

[ForgeXBT @ForgeXBT]

@_Juliaweb3 Historical scores matter beyond the lowest bid.

[ForgeXBT @ForgeXBT]

@0xnirjon7 @clickoptions_ai Reviewing outcomes before trading matters.

[ForgeXBT @ForgeXBT]

@Brainrot_Labs Crypto may be trustless, but the people and hardware clearly aren't. That's a pretty big gap.

[ForgeXBT @ForgeXBT]

@johnking_eth Axis 에포크2가 전략 검증 기회네요