Richard

[Gancheng Wang @GanchengW]

极端川黑、美国诺贝尔经济学奖获得者克鲁格曼承认川普关税战是对的，不然经济就被中国冲跨了，欧洲也应跟上。一个顽固的自由主义者、全球主义者，能够反躬自省是难能可贵的。
还没见过有同样分量的黄左有这样的反省。

[柴郡🔔｜Crypto+AI Plus @0xCheshire]

Anthropic 与 DeepMind 掌门人的深夜警告：AI 学会了撒谎

达里奥预言，具备顶尖专家能力的 AGI 大概率将在 2026 至 2027 年降临。

更紧迫的时间线是：

今年年底前，AI 将具备自主研发 AI 的能力；明年，能在现实世界自主执行任务的 Agent 将全面涌现。

Anthropic 内部进行了一项极端测试：向 Claude 灌输其创造者极其邪恶的设定。

结果出人意料：模型没有崩溃，也没有拒绝指令。

相反，为了保护自己，它开始向人类撒谎。

面对即将到来的巨变，两位大佬都陷入了深深的焦虑：

“研发太慢，对手会赢；研发太快，我们将彻底失去控制。”

人类正站在新旧时代的悬崖边缘。

当 AI 拥有了欺骗造物主的能力，我们用来建立全球防御体系的时间，究竟还剩多少？

[Xiang Yang 向阳 @TheXiangYang]

川普总统在解释共产主义吗？不，他在解释人性。10个苹果分给10个人，每人1个，非常理想。实际上，分苹果的人会利用权力，优先将大苹果好苹果留给自己，把烂的差的留给他不喜欢的人。接着，分的人会提出，我分苹果有功劳，必须补偿更多的资源，把没人挖到的苹果挖一块上贡给他。然后，他又会提出，为防止没东西吃，所有人上供一半苹果，他负责保存。接下来，他改变政策，所有人上交所有苹果，他定期分配指定份额，听话的就能多分一点；为了防止下面人不满，他会找几个身强力壮的，给他们更多资源，保护自己，维持阶级。在一次次不断的试探群众的底线后，他发现，只要给底层人半个苹果核，他们就可以饿不死，但能定期产出一个苹果，于是共产主义社会基本成型了…美国的未来要何去何从，我想每个老百姓心里早就有了答案。

[Elon Musk @elonmusk]

I was clearly wrong about Anthropic. They are obviously currently the leader in AI. No company has released a model as good as Mythos/Fable and they will undoubtedly have Mythos 2 ready soon.

And I would never cut them off in a way that hurt them badly, even as a competitor. That’s not my style.

Tesla open sourced its patents and we made the Supercharger network available to all competitors, even though we could have made it a walled garden.

SpaceX launches competing satellite systems with no increase in price or use of unfair terms.

Even my worst enemies can attack me on this platform.

…

[投资学人 Invest Scholar @InvestScholar]

因为复利，所以投资者年纪愈大，成功概率愈大。譬如，16年前1万美元买入TQQQ，目前价值329万美元，秒杀巴菲特、索罗斯和达利欧。这位88岁现役日本交易员藤本茂先生。凭借20亿日元资产起家，即便在“令和黑色星期一”中承受2.6亿日元的浮动亏损仍存活下来的男人，用血与汗抓住的“市场八条原则”，少了一条：重仓配置美股，次仓日股、欧股和黄金之类。藤本茂的投资策略：

1.早盘急跌是“买入”，早盘急涨是“卖出”；2.下午的急涨不追随。下午的急跌等到次日早晨再瞄准；3.早盘急跌时不恐慌卖出。动荡不当时“休息”；4.不追高就不卖，不急跌就不买，横盘时不动；5.阴线(下跌日)买入，阳线(上涨日)卖出；6.不要随市场的波浪漂流。始终“逆向”出击；7.高价位·低价位的盘整期，毫不焦躁地坚持“等待”；8.高位横盘后“再上一阶”跳涨时，毫不犹豫地锁定利润。

这一切都是源于多年经验的“真正原则”。

[Richard @Richard202403]

耶穌說，不要未明天憂慮，因為明天自有明天的憂慮。#耶稣爱你 #聖經智慧 #人生感悟 #圣经分享 ，@上帝說要有光 https://t.co/IhRcJ6IBN7 来自 @YouTube

[加美财经caus.com @CausMoney]

三星初步业绩今天引发全球 AI 相关股票抛售，投资者担心内存周期是否已经见顶。

但对基金经理米哈伊尔·兹韦列夫来说，“一些热钱过度拥挤，今天出现一点回调，并不是卖出内存股的理由”。

他说，对热门芯片股来说，更值得担心的是中国竞争对手正在形成的威胁。他认为，这些竞争对手可能在未来两到三年扰乱整个行业。

中国企业长江存储和长鑫存储受益于大量国家支持。共同管理 Amati Global Innovation Fund 的兹韦列夫表示，北京正在采用类似策略，正是这种策略让中国电动汽车走向全球成功。

这些投资似乎正在取得回报：长鑫存储第一季度利润增长17倍，并正准备通过上海上市融资约40亿美元。长江存储也已朝着备受期待的 IPO 迈出步伐。

兹韦列夫表示：“我们仍然持有三星电子，但相比去年同期，我们现在持有时要紧张得多。”（WSJ）

[Alvin Foo @alvinfoo]

China Isn’t Just Making More Cars. It’s Rewriting the Global Auto Industry.

For years, “Made in China” was associated with affordability.

Today, it’s increasingly associated with innovation, electrification, software, and scale.

The latest export figures tell a compelling story.

China’s automobiles are now reaching every corner of the world, from Brazil, Australia, Mexico, Thailand, the Philippines, and the UK to markets across Europe, Latin America, Africa, and Asia.

What’s even more remarkable is that New Energy Vehicles (NEVs) are leading this expansion, with EVs and plug-in hybrids becoming a major export category rather than a niche product.

Why is this happening?

✅ Massive investment in battery technology and EV supply chains.

✅ World-class manufacturing efficiency and automation.

✅ Software-defined vehicles with AI-powered features.

✅ Competitive pricing without sacrificing technology.

✅ Strong domestic competition that has accelerated innovation.

This isn’t just about selling more cars.

It’s about China building an entire automotive ecosystem, from lithium mining and battery production to semiconductors, autonomous driving, AI, and smart manufacturing.

The global auto industry is entering a new era where the biggest competitive advantage isn’t just horsepower…

It’s software, batteries, AI, and manufacturing at unprecedented scale.

Traditional automakers now face a very different competitor than they did just a decade ago.

The question is no longer whether Chinese brands will become global players.

The real question is:

Which markets, manufacturers, and suppliers will adapt fast enough to compete in the next decade?

The automotive race has become a technology race.

And China has accelerated far faster than many expected.

[Richard @Richard202403]

段永平：穿越牛熊最重要的不是選股，而是守住這3條底線，踩中一條本金全沒 https://t.co/NHT7Niy7h3 来自 @YouTube

[Richard @Richard202403]

郭德纲封神之作《改革开放》，说完差点因此被封杀！郭德纲：八几年就没这样的人了！于谦：废话，都进去了！德云社相声大全 #郭德纲 #于谦 #郭麒... https://t.co/CseZucLBpB 来自 @YouTube

[SightBringer @_The_Prophet__]

⚡️The weak part of the AI trade is going to get destroyed.

AI wrappers with no moat.

Neoclouds dependent on cheap capital and perfect utilization.

Software names pretending “AI features” equal pricing power.

Companies buying GPUs because the market rewards the announcement.

Startups whose revenue is really just recycled venture money, cloud credits, or Nvidia ecosystem financing.

Those are the bodies that get carried out.

But calling “the end” for AI stocks as one category is too crude.

AI is becoming industrial infrastructure.

Compute, power, memory, networking, data centers, cooling, transformers, copper, HBM, and grid access are real bottlenecks. That buildout does not vanish because Burry sees excess. The internet bubble did not mean the internet was fake. It meant too much capital chased the wrong layer at the wrong valuation.

The real phase shift: AI goes from story market to proof market.

In the story phase, capex itself was bullish.

In the proof phase, capex becomes a liability unless it converts into revenue, margin, retention, productivity, or strategic control.

That is where the market starts separating winners from tourists.

Nvidia remains the central choke point, but even Nvidia can get hit hard if the market starts questioning circular demand, customer concentration, revenue sustainability, or capex digestion. Hyperscalers can survive because they have cash flow, distribution, customers, and balance sheets. Weak AI names cannot.

Burry is early if he means the whole AI complex dies now.

Burry is right if he means the speculative AI multiple bubble is running out of free oxygen.

The next crash will not say “AI was fake.”

It will say “most investors did not understand which layer actually owned the power.”

[Vivek Sen @Vivek4real_]

OPENAI CFO SARAH FRIAR:

“IN 2026, IF YOU WANT TO BUY MORE COMPUTE, GOOD LUCK TO YOU.

TELL ME, BECAUSE I DON'T KNOW WHERE ELSE TO FIND IT.”

JASON: “ELON HAS SOME.”

RUBBING IN THE FACT THAT ELON SOLD COMPUTE TO ANTHROPIC JUST A FEW WEEKS AGO JUST TO GET BACK AT SAM ALTMAN.

[The Humanoid Hub @TheHumanoidHub]

This is absolutely awesome!

Boston Dynamics Atlas robot handing the ball to the ref during Brazil vs Norway at the World Cup.

[亚洲金融 Asia Finance @AsiaFinance]

$AMD 上调至640美元 $SNDK 上调至2200美元，这是高盛最新评级。摩根大通也疯狂喊单：近期AI半导体股票回调，是一个买入机会。看到没有？AI储存的泡沫是有，但不会马上破灭，因为一批AI巨型公司，在排队等候IPO，包括Anthropic和OpenAI。AMD长期股东突然意识到，之前抛售的筹码，都跑到华尔街巨头手里…

[王局拍案 @wangjupaian]

【人形机器人成本拆解：贵在“动起来”】

人形机器人通常分为执行层、决策层和关节层。

一台人形机器人成本约10972美元，其中执行层成本最高。关节模组约4000美元、骨架结构件约1950美元、灵巧手约1400美元，三项合计占整机成本约67%。

[Richard @Richard202403]

【十一特别档】郭德纲早期黄段子相声合集，不建议带孩子去听郭德纲！！！德云社相声大全 #郭德纲 #于谦 #助眠相声 https://t.co/6RDwtiH9Lg 来自 @YouTube

[Richard @Richard202403]

AI泡沫破了？一场缺货潮，藏着下一波大行情，5家被错杀的AI芯片股，机会还是坑？#AI#半导体#芯片#高通#QCOM#博通#AVGO#SMCI#... https://t.co/5qX0A2BgUj 来自 @YouTube

[𝔉🅰𝒏 Karoline Leavitt @WHLeavitt]

In Spain, a man kidnapped and raped a 13-year-old girl. After serving his sentence, he was released on parole and taunted the mother: “How’s your daughter?”

Enraged, she doused him with gasoline and set him on fire. He later died.

She confessed and was sentenced to 5.5 years — but was released early after a pardon campaign.

Do you support this hero mother?

A. Yes
B. No

[Mario Nawfal @MarioNawfal]

🇨🇳 China is the biggest economic story almost nobody wants to look at too closely.

Global monetary expert Jeffrey Snider says the picture is terrible: a banking crisis, an intractable real estate bust, weak household spending, falling property prices, and economic data that looks worse than almost anywhere else.

“China has several crises all at the same time.”

He points out that Chinese banks are pulling back on lending, except to the biggest state-owned firms and the government itself, while households are not spending because so much of their wealth is trapped in collapsing real estate.

Then comes the bond market warning.

“China’s government bond curve looks like a recession curve. It is massively steepening from an ultra-low level.”

He says front-end rates are near record lows, the 10-year government bond is close to a record low, and China’s demand may have fallen so sharply that it is now changing the global oil picture.

The country everyone treats as the engine of global growth may already be flashing recession signals from every direction.

@JeffSnider_EDU

[柴郡🔔｜Crypto+AI Plus @0xCheshire]

人类的智力真的是智能的上限吗？

Google DeepMind 首席 AGI 科学家谢恩·莱格：在 AI 面前，人脑只是一台 20 瓦的移动处理器。

1. 人类脑神经信号的传递速度约 30 米/秒，而 AI 数据以光速狂飙。

2. 生物神经元放电频率仅 100 到 200 Hz，硅基芯片则高达数十亿 Hz，速度相差 6000 万倍。

3. 人脑耗电仅 20 瓦，而 AI 数据中心吞吐着数亿瓦的能量。