CryptoJuan0x77

[huaijiang @huaijiangzhu]

my spicy theory is that chinese ai labs keep winning because of culture, not talent or resources.

chinese labs still have a deeply hands-on engineering culture. deepseek and kimi are flat organizations where science and engineering are fused: the same people move between algorithms, data, and infra, doing whatever it takes to make the model work.

but sf tech bros have decided that “researcher” is the high-status title while infra is merely support work. every new sf ai startup calls itself a “lab,” every ambitious engineer quietly upgrades their title to “research engineer,” and the infra work is left to whoever failed to escape it.

but at frontier scale, infra determines experiment velocity, and experiment velocity determines research output. at frontier scale, infra **is** research.

[CryptoJuan0x77 @juan0x77]

@KostasSiagas @CryptoLibraryGR Nice work x

[CryptoJuan0x77 @juan0x77]

Starting my journey 🥷

[nftcrypto3.eth 🍌 @nftcrypto3]

最值得关注的 10 个中文KOL观点（https://t.co/aF4t9Vwrds 总结）2026.8.20

1. 憨巴龙王 (@dotyyds1234)【方向交易看人性，套利看智力】

核心观点： 憨巴龙王认为，二级方向交易与套利考验的是两套完全不同的能力。���向交易更考验人性与风险控制，因此他更愿意参考大资金的判断���套利则更偏结构理解、计算和执行，小资金交易者同样可能提供高质量思路。

延伸解读： 在他看来，大资金的优势不在于“更聪明”，而在于更重视赔率和风险——错的时候少亏，对的时候敢于放大利润。相反，不少套利交易者虽然擅长发现结构性机会，却未必适合承担高杠杆方向风险；会做套利和会做趋势，本质上是两种不同的交易能力。

原文：https://t.co/thtXCPyH4W

2. qinbafrank (@qinbafrank)【SEC 监管范式转向】

核心观点： qinbafrank 认为，本轮币市反弹除了流动性改善，更重要的变量是美国监管方向出现变化——SEC 正从“执法式监管”逐步转向“规则式监管”。其中，他尤其关注 SEC 提案对“代币”与“投资合同”的区分。

延伸解读： 如果这一框架最终落地，关键意义不只是减少监管不确定性，而是可能为美国重新打开一条合规的公开代币融资路径。对加密市场而言，这意味着一级发行、交易平台和项目融资都有机会获得更清晰的制度边界。

原文：https://t.co/jEZRJfu4gI

3. Pike时政财经分析 (@bigbirdflychan)【AI 开始兑现生物医药】

核心观点： Pike 将 Moderna 股价大涨与 AI 驱动的个体化 mRNA 癌症疫苗进展联系起来，认为这不只是一次新药突破，也是 AI 在真实临床场景中的一次重要兑现。

延伸解读： 他的核心判断是，AI 的价值正在从模型、软件和算力逐渐向医药等实体产业外溢。如果 AI 能持续缩短药物研发周期、提高患者筛选和个体化治疗效率，市场对 AI 的定价逻辑也可能从“卖铲子”扩展到下游应用。

原文：https://t.co/Waf11rSXwT

4. Feng Liu (@fishkiller)【Stripe 押注 Agent 金融】

核心观点： Feng Liu 认为，Stripe 近期一系列激进收购并非单纯扩张支付业务，而是在围绕稳定币与 Agent Economy 提前搭建 AI 时代的金融基础设施。他把其中的核心概念概括为“智能资本”。

延伸解读： 在这一框架下，支付公司的角色会从“帮企业收钱”进一步升级为“帮助软件和 Agent 持有、调用和分配资本”。如果 Agent 真正成为经济主体，稳定币、钱包、结算和资金调度都可能成为新的基础设施层。

原文：https://t.co/hHWIaH7Oof

5. taresky (@taresky)【EtherFi 自动借贷风险】

核心观点： taresky 对 EtherFi 的借款功能发出风险提醒：他指出，部分用户的 USDC/USDT 余额可能在没有充分感知的情况下被转入 Aave 并作为抵押品使用，由此叠加额外的协议与清算风险。

延伸解读： 他质疑这种设计带来的额外收益是否足以覆盖新增风险，并建议用户检查并关闭不需要的借款功能。对 DeFi 用户来说，这也再次说明“余额显示在哪里”和“资产实际上被部署到哪里”并不总是一回事。

原文：https://t.co/O7N0fHjYi8

6. 蓝狐 (@lanhubiji)【币市上涨的三重驱动】

核心观点： 蓝狐将近期加密市场上涨归因于三条主线：美国财政部增加长期国债回购、SEC 提出新的加密发行规则，以及白��与加密行业重要人物会面。

延伸解读： 这三件事分别对应流动性、监管预期和政治信号。蓝狐的框架是：当宏观流动性改善，同时监管尾部风险下降，市场风险偏好就更容易被重新激活。

原文：https://t.co/NazK24U8vI

7. 大老师Bugsbunny｜DRAM UP only (@octopusycc)【算力正在变成金融资产】

核心观点： 大老师认为，AI 算力正在从单纯的技术基础设施演变为一种可被机构配置和金融化的资产。NVIDIA 与大型资本合作引入第三方资金，本质上是在为 AI 基础设施建立新的资本形成机制。

延伸解读： 他特别提到 Falcon Finance 将算力资产 RWA 化，再通过 Uniswap 等链上市场提供流动性，并进一步作为抵押品铸造 USDf。核心逻辑是把“算力产生的未来现金流”变成可交易、可抵押、可融资的链上资产。

原文：https://t.co/4QOpmpJiWl

8. 鸭哥 (@grapeot)【AI 下一瓶颈是行为���据】

核心观点： 鸭哥认为，AI 公司下一阶段争夺的关键资产之一，是“人类在真实工作过程中留下的行为数据”。他总结了四种主要获取路径：收购成熟产品、购买破产企业数据、通过 SaaS 合同获取，以及自建数据闭环。

延伸解读： 他的判断是，数据飞轮已经不再只是一个模型问题，而是一整套资本、协议、治理和产品工程能力。模型能力逐渐趋同后，谁能持续获得高质量、真实且可授权的行为数据，可能成为新的竞争壁垒。

原文：https://t.co/zbzVceBsWo

9. 憨巴龙王 (@dotyyds1234)【宇树定价或被情绪推高】

核心观点： 憨巴龙王认为，宇树的短期定价可能同时受到两层因素推高：一是 A 股无法直接做空，部分币圈资金存在信息差优势；二是市场容易把宇树与长鑫等热门叙事混在一起，从而进一步放大估值。他因此判断，宇树本身估值已经不低，再叠加情绪���价后，风险收益比并不理想。

延伸解读： 不过他也承认，市场短期可以持续非理性，价格完全可能再冲 2–3 天、上涨约 20%。因此他没有大规模押注，只在 Hyperliquid 实盘做空约 30 万美元，体现的是“看空估值，但尊重短期情绪和路径风险”的交易思路。

原文：https://t.co/v8ckDThjzg

10. yourQuantGuy (@yourquantguy)【Hyperliquid 生态新机会】

核心观点： yourQuantGuy 关注 Hyperliquid 生态中的新交易平台 Entropy，重点提到其潜在空投、机构投资背景以及 100% 反佣机制。他同时提到 Boros 的 P2P 功能和 crossEx 套利机会。

延伸解读： 他的关注点本质上是新市场启动阶段的“流动性红利”：空投激励、手续费返还、P2P 做市和跨平台价差往往会同时出现。不过这类机会高度依赖成交深度、手续费结构和激励持续性，需要区分真实可套利空间与单纯的早期补贴。

原文：https://t.co/nodISvwOMd

[Miyamoto @iruletrenches]

if u got 10k, you’re just two 10x away from 1mil

crazy, right?

now lock tf in

quality > quantity. stop trading slop and getting rinsed by kols, focus your time on doing real DD and finding the only two right picks you need

[深情丨先信🔶BNB @Web3shenqing]

Crypto changed my life and it can change yours too.

[CryptoJuan0x77 @juan0x77]

Memes are getting crazy we are printing 🥷💸

[CryptoJuan0x77 @juan0x77]

BITCOIN IS GOING PARABOLIC 🚀
BEARS ARE GETTING REKT!
ARE WE GOING PARABOLIC ONCE MORE?

[bobby @bigbagsbobby]

the best advice i can give if you want to actually last in memecoins is this: build your own intuition.

borrowed conviction can make you money. but it puts a ceiling on your upside.

the biggest shift for me was when i started understanding why certain coins would run. once i knew the why, i knew what to look for, and i stopped relying on others to tell me what to buy.

being able to see a coin and just know its a cook is a superpower. this pivot was also the moment i started being consistently profitable.

front-running KOLs, wallet tracking, chasing volume, etc are all viable strategies. and i'm not saying to stop taking those trades if they're working.

but, when you start forming your own thesis, and have your own conviction that a narrative is undervalued, is when you'll start finding outsized returns. and when you can actually hold long enough to see them play out.

i found $USDUC below 200k, and was able to see the vision for a 400x+.. that was not luck.

building this sixth sense takes time. hours in the trenches, months of watching runners, making mistakes, and adapting.

it doesn’t come fast, and it doesn’t stay static. the market changes constantly, and you have to keep up.

but if you never learn to think for yourself, or at least put constant effort into developing this skill, your potential will stay capped.

the practical side of this is simple: stop gambling, study runners, journal your trades, hunt for undervalued narratives, and fail a lot.

do this long enough and your intuition will sharpen, things will start clicking, and you'll get consistent... the money will also start coming in.

or you can keep gambling on 70 shitters every hour, idrc.

unstable coin.

[Elie Habib @eliehabib]

Hot take: AI won’t cheapen human work. It’ll nuke the middle and split the market for thought in two.

🔻One side: infinite output. Drafts, code, decks, analysis, strategy docs. A teenager with Claude and Codex now ships what used to need a staffed team and three weeks of meetings.

🔺Other side: judgment. Knowing what’s true. What’s noise. What breaks at scale. What compounds. What’s a hallucination wrapped in confidence.

The first market is collapsing into abundance. The second is becoming a scarcity market.

When everyone can generate, the only thing that matters is taste, discernment, and the ability to say: “this is the one worth betting on.”

AI makes production cheap.
Conviction gets expensive.

[Jenny Zhang @jennyzhangzt]

Introducing Hyperagents: an AI system that not only improves at solving tasks, but also improves how it improves itself.

The Darwin Gödel Machine (DGM) demonstrated that open-ended self-improvement is possible by iteratively generating and evaluating improved agents, yet it relies on a key assumption: that improvements in task performance (e.g., coding ability) translate into improvements in the self-improvement process itself. This alignment holds in coding, where both evaluation and modification are expressed in the same domain, but breaks down more generally. As a result, prior systems remain constrained by fixed, handcrafted meta-level procedures that do not themselves evolve.

We introduce Hyperagents – self-referential agents that can modify both their task-solving behavior and the process that generates future improvements. This enables what we call metacognitive self-modification: learning not just to perform better, but to improve at improving.

We instantiate this framework as DGM-Hyperagents (DGM-H), an extension of the DGM in which both task-solving behavior and the self-improvement procedure are editable and subject to evolution. Across diverse domains (coding, paper review, robotics reward design, and Olympiad-level math solution grading), hyperagents enable continuous performance improvements over time and outperform baselines without self-improvement or open-ended exploration, as well as prior self-improving systems (including DGM). DGM-H also improves the process by which new agents are generated (e.g. persistent memory, performance tracking), and these meta-level improvements transfer across domains and accumulate across runs.

This work was done during my internship at Meta (@AIatMeta), in collaboration with Bingchen Zhao (@BingchenZhao), Wannan Yang (@winnieyangwn), Jakob Foerster (@j_foerst), Jeff Clune (@jeffclune), Minqi Jiang (@MinqiJiang), Sam Devlin (@smdvln), and Tatiana Shavrina (@rybolos).

[CyrilXBT @cyrilXBT]

A Russian developer just figured out how to learn anything 10x faster.

He saved 1,460 hours using 3 tools most people already have.

NotebookLM + Gemini + Obsidian.

Here is how it works.

You dump any source into the system.

YouTube videos. Articles. Courses. Podcasts. All of it.

The AI strips out every duplicate concept across all your sources.

Then it keeps only the information you do not know yet.

Think about what that means.

20 YouTube videos on the same topic each one repeating the same 20% of information you already know.

This system removes that 80% of noise automatically.

What used to take a month of learning compresses into 15 minutes of pure signal.

No repetition.

No fluff.

Just the exact knowledge gap between where you are and where you want to be.

This is not a productivity hack.

This is a completely different way of consuming information.

The people who build this system this week are going to compound their knowledge at a rate that is genuinely unfair compared to everyone still watching the same tutorial for the third time.

Full credit to @zproger on YouTube.

Go watch his breakdown.

Then build the system today.

[CryptoJuan0x77 @juan0x77]

@SolportTom I log into Bullx and Axiom and no fucking motion from https://t.co/VcwdJRLR4A, that is completely sad! U are a piece of shit Tom

[tim. @ygg0f]

lock in. go deep.

[CryptoJuan0x77 @juan0x77]

@ryanYELLAND @DegeCoin_ @grok what percent of the liquidity is locked ? ca: CPG7gjcjcdZGHE5EJ6LoAL4xqZtNFeWEXXmtkYjAoVaF

[CryptoJuan0x77 @juan0x77]

@ryanYELLAND @DegeCoin_ It looks like only about 66% of DegeCoin’s liquidity is currently locked, according to RugCheck – meaning roughly one-third is unlocked and active! That's why it shows unlocked! Its not fully unlocked though!

[CryptoJuan0x77 @juan0x77]

@bos_ot04 @DegeCoin_ It looks like only about 66% of DegeCoin’s liquidity is currently locked, according to RugCheck – meaning roughly one-third is unlocked and active