DZ

OUTSIDL

[Cointelegraph @Cointelegraph]

🔥 LATEST: Banks and regulators across Europe, the Middle East and Asia are testing quantum-resistant crypto wallets and onchain transfers using a NEAR testnet.
https://t.co/eSNAKb2fEr

[Phyrex @PhyrexNi]

周末和预想中的一样顺利度过，但我觉得下周有要开始紧张起来了，最重要的应该是北京时间 27 日凌晨英伟达的财报了，现在很多小伙伴都重仓 AI 和半导体，而英伟达的财报基本就是这个领域的探照灯，英伟达的财报不但要好，还要大幅的超过预期，AI 赛道才会继续有大量资金买单。

在英伟达的财报前一天还有美联储最关注的 PCE 数据。现在就业已经开始走弱，虽然市场预期 七月 的 PCE 还是不错的，但如果油价仍然走高，那么七月的好数据几乎就没意义了，八月的数据通胀可能会继续提升，到时候因为劳动力降低带来的加息预期减弱可能又会被提起来了。

紧接着两天又是杰克逊霍尔全球央行年会，这也是沃什成为美联储主席以后第一次参会。最近长期美债收益率已经明显上升，30 年期一度到了 2007 年以来最高的位置，半导体指数上周也跌了大约 5%，所以这次市场会非常关注沃什对通胀、经济增长以及利率路径的态度。

最后就是 Bitcoin ，虽然周末有过两次大的波动，但还是顺利稳定在 77,000 美元左右，而周一晚上开始流动性就开始回归了，这时候 bitcoin:native 的走势会更加明显，这次的反弹能不能冲破 80,000 美元，我觉得周一就能看出一点迹象，主要是看成交量，资金和 ETF ，顺便看看 $MSTR 会有什么动作。

@Gate Crypto、美股、港股、韩股、黄金、CFD、预测市场一站交易

[DZ @wanglei198888]

@urclavicle @intel @IronClawAI 保持客观

[DZ @wanglei198888]

@_kuti_Nelson 有迹可循，稳定币供应的突然增加和消失，似乎是数据失真，但可能被利用来炒作

[DZ @wanglei198888]

开创

[BSCN @BSCNews]

How Does NEAR Let You Pay for AI With Staked Tokens?

On July 30, NEAR (@NEARProtocol) introduced a new way to pay for AI inference and always-on agents.

Instead of paying for an AI subscription with a credit card, users can stake $NEAR and receive monthly compute credits.

The important part?

The NEAR tokens are not spent.

They remain staked until the user decides to unstake them, at which point the tokens return to the user's wallet.

Here’s how the system works:

(1) Stake $NEAR

Users lock NEAR tokens into the system.

The amount staked determines how many compute credits they receive for NEAR AI services.

It works more like a refundable deposit than a traditional subscription.

(2) Use the credits for AI

Those credits can be used across all 43 models currently available through NEAR AI.

That includes models from major AI companies such as Anthropic, OpenAI and Google.

Users can therefore access AI computation without maintaining a traditional credit card subscription.

(3) Adjust the stake when needed

The system is designed to move with usage.

Users can increase their stake when they need more computing capacity or reduce it when their requirements fall.

They can also lock additional NEAR to purchase one-off credits for short-term projects.

(4) Unstake and recover the tokens

When users no longer need the service, they can cancel the arrangement and unstake their NEAR.

The tokens are returned rather than being consumed as payment.

The capital backs access to computing instead of simply being transferred to a service provider.

So why is NEAR doing this?

The bigger idea goes beyond AI subscriptions.

NEAR has been exploring what happens when software agents become economic participants.

AI agents may eventually need to pay for computing, data, storage and other services without relying on human-operated bank accounts or credit cards.

NEAR's argument is that the token securing the network could also become part of that machine economy.

In this model, NEAR is not simply a token used for staking.

It can also act as a claim on the network's available computational resources.

There is also a tokenomics angle...

When users stake NEAR to back an active AI subscription, those tokens are temporarily removed from circulating supply.

A single subscription would have little impact.

But if AI inference and autonomous agents generate sustained demand, more NEAR could remain locked to support that activity.

That creates a link between actual network usage and the amount of NEAR temporarily taken out of circulation.

The important distinction is that this demand would come from usage rather than speculation.

More AI activity could therefore mean more NEAR being locked at any given time.

The real test will be whether developers and AI agents actually create enough demand to make this model economically meaningful.

[DZ @wanglei198888]

@zacodil 欢迎回归Near风格

[DZ @wanglei198888]

@zacodil 机构采用会极大促进影响力

[DZ @wanglei198888]

@aescobar0g ref和burrow还没上线的时候就坚定支持 $NEAR .藐视波动，从不怀疑！

[DZ @wanglei198888]

👀

[RWA Foundation @RWAFoundation_]

BlackRock (@BlackRock) where the $ value is spread… 👀

[Vadim (AI, ⋈) @zacodil]

Most "AI tokens" don't get you any AI. They get you emissions and a narrative. NEAR just became the exception.

Plainly: you stake NEAR and it turns into AI you actually use. Confidential inference and hosted agents running Claude, GPT, Gemini and open models. You keep the NEAR and unstake whenever you want.

Compare that to TAO, the token everyone holds up as the AI trade. You stake TAO to earn emissions, and what comes back is more TAO. It's yield farming with an AI network bolted on. Staking it doesn't get you a single inference call.

That's the whole difference, and it's not subtle. One token pays you in more of itself. The other pays for real compute and gives your capital back. For anyone building a product or running research, only one of those is worth anything.

NEAR pays for using AI. TAO pays you for holding TAO. Only one runs your workload.

[Justin Bons @Justin_Bons]

Ethereum promised sharding for years. It finally shipped: It's called NEAR!

[DZ @wanglei198888]

@zacodil 需要黄仁勋级别的人做背书

[Sal Ternullo @sal_ternullo]

A big part of what we do at SVRN is finding new ways to commercialize the NEAR product set. This collaboration with @Oracle is a great example, where we piloted an integration of @near_intents with their point of sale system Simphony.

[NEAR Protocol @NEARProtocol]

MAINNET UPGRADE

Quantum-safe signing is now live on NEAR mainnet.

This makes NEAR one of the first blockchains with a NIST-approved post-quantum signature scheme in production.

The upgrade also launches dynamic resharding, so NEAR Protocol can scale itself automatically 🧵

[SanTi Li @SantiLi1021]

https://t.co/RwvTgQgoMw

[DZ @wanglei198888]

@AllOut2027 完全同意

[DZ @wanglei198888]

@y_cryptoanalyst @zama @NEARProtocol @randhindi 这么懂技术，精神这么高洁，怎么对HL和SUI之类劣迹斑斑的协议如此甘之若贻的？我也挺好奇这事儿。基本逻辑，本来觉得zama好像是有点儿极客精神，但要这么收集kol搞拉踩，那好像是有点嫉妒在的，属于自暴自弃了

[DZ @wanglei198888]

@y_cryptoanalyst @zama @NEARProtocol @randhindi 你只不过是搞了个商务合作，跟认知是两个层次的事，如果你能实现一个智能合约的合规模块，并且开源，那我还一如既往认可你的技术大神人设，要不你就只是个接单的，怎么样？🤣

[DZ @wanglei198888]

@y_cryptoanalyst @zama @NEARProtocol @randhindi 作为有市场追求的团队，Zama未来能否保持全部开源现在也言犹过早

[DZ @wanglei198888]

@y_cryptoanalyst @zama @NEARProtocol @randhindi 👍说的很好，那闭源的该以什么作为闭源的借口呢🤔？当然开源属于道德高地，没人反对是一种精神，但开源智能合约的已知能力不能全部兑现现实需求的时候该怎么办呢？另外，开源并不只代表高尚，也代表推责，一体两面