狗卷棘

[Zora @Zorathzzz]

2 weeks ago: $20k. now: $600k.

[徐可KeXu @xukego]

喜欢哪种尺寸🧩

[狗卷棘 @femtoo_griffith]

@woaitaoershi 都他妈装逼犯 说几把英文

[Colin Wu @colinwu]

深刻地感觉到

今天会用 GPT/Claude 的人，绝大多数，和只会用豆包的人相比，尤其是小孩（或他们的父母），人生未来的发展将不是一个阶级，或者可能本来就不是一个阶级。

当然还有更糟糕的，豆包都不会用，还在自以为是、用自己或环境中愚昧的知识教育小孩。

[Tibo @thsottiaux]

Let ChatGPT *work* for you. How many time have you wanted to negotiate your internet bill, get rid of all those spam emails you’re subscribed to, or find the perfect deal for something you wanted to do or buy. It’s quite literally just one prompt away, all from the comfort of your phone. It does at least 20 things for me every single day and I’m still surprised.

[VP @VibesPatrol]

victory does not start on the battlefield, in the gym, or at the desk. it starts in the mind long before the first move is made.

stop waiting for proofs to believe. the ones who go furthest reverse the order. they see the outcome with such clarity that their nervous system begins treating it as already real. the body follows. the actions align. the world rearranges itself around the image they refuse to release.

imagine where you will be, and it will be so.

not as wishful thinking. as a directive. the mind that can hold a future with total conviction is already halfway there. the rest is just the work of closing the gap.

see it first. live from it. then let reality around you catch up.

[girass0lle @emllysz]

quando voce quer fumar menos mas eh assim q a desgraca da fumaca te olha

[X Freeze @XFreeze]

"A collapsing birth rate is the biggest danger civilization faces by far

My advice regarding starting a family is start immediately… Just have kids. You won't be sorry and it will be the best decision you ever make"

— Elon Musk

[🧬Maxpein🧬 @maximumpain333]

THE FIELD OF COSMIC LOVE AND UNIVERSAL LIFE

The Primordial Field is the primary and structural force of space, within which, forms are gestated, from the atom to the galaxy, from the planets to our bodies.

This field gives rise to the toroid, the primordial electromagnetic egg of life, the vital flow and reflux that nourishes from the core outwards, producing the matrix and pattern, in which all things are inscribed.

Among the different forms created, there exists a field resonance, an essential identity that links and vivifies them, a fundamental unit that is their reason for being and their essence.

This primordial field is the expression of Cosmic Love, Universal Life, Divine Consciousness, transcendental beauty, spontaneously creating a sacred game.

Knowing this field and its functions was the forbidden fruit...

✨🙌🏾💫

[CoinMarketCap @CoinMarketCap]

LATEST: 🤖 Elon Musk says AI will likely be in control in 10 years: "If the difference in intelligence between AI and humans is vastly greater than between AI and chimpanzees, it's hard to imagine the chimpanzees would be in charge."

[AnirbanBandyopadhyay @anirbanbandyo]

Graph mode, graph engineering, graph abstraction, graph of nodes, graph of contexts, and then comes morphogenesis, that is life, that is birth of consciousness.

[A.J. @aj_dev_smith]

opus 5's digital plants can flower and fruit now. plus we have new species!

every petal, every seed, every vein is one hormone diffusing between cells. still not a single shape anywhere in the code. just a ton of math and chemistry

[Mathelirium @mathelirium]

The coolest account on @X by far!

No one comes close!🫡

[kuno @kunoo]

“I don’t gamble on Crypto, it’s too risky”

Your grandpa when he was your age:

[gostosa, chapada & consciente. @nahbrisaaaa]

a sociedade: maconheiro é tudo vagabundo preguiçoso e jamais vai ter sucesso na vida

o maconheiro:

[Blxckie🥷🏽 @Blxckie1__]

You do get over it on a random day.

[caro 猫猫头🐈‍⬛ @lin_yiliu]

看了无限游戏这本书
ai 时代最好的解法就是无限游戏
人不能只有1-2个工作 应该有5-6个工作
或者不能叫工作 叫有意义且能带来回报的事情
否则太多东西都被坍缩到一个纬度比较了
工作这个比较的维度以后要消失了
别的维度 就会崛起

总会有新的标签和时代的符号来定义这个事情

[Rocky @Rocky_Bitcoin]

这周末约中大历史系的博士一起喝茶，正好聊到最近国家严查税务的问题，代入历史角度的观点，还蛮有意思，让我娓娓道来！

他说，读历史读多了，研究历史研究多了，你会发现一件很诡异的事，历史如同钟摆，一刻不停，变的是外表，但内核从来未曾改变。

大厦倾塌之前，从来没有人会敲锣打鼓通知你。但如果你足够冷静，会发现历史总是习惯性地提前给你塞三个信号，不早不晚，精准地出现在每一个转折点的前夜。

第一个信号：财产权开始变得不稳定。你以为是自己的东西，不再确定明天还是不是你的。

第二个信号：税收政策开始"带情绪"了。不再是规矩框住的预算逻辑，而是应激的、惩罚性的，甚至是清算性的。

第三信号：资金的流动性被死死按住了。钱宁愿烂在地里、躲进夹层、跨境消失，也不敢在市面上流转。

假如只是单个信号出现，咬牙还能扛。但这"三合一"的组合一旦凑齐，历史的账，就进入最后的清算程序了，这时候尤其要留心。

然后，他按照上述思路，我们一起探讨起了历史，历史为镜，看看这逻辑是否恰当。

第一笔账：明朝崇祯时期，一个帝国的陨落

很多人把明朝灭亡的锅扣在李自成头上，其实真要掰开来看，李自成不过是压垮骆驼的最后一根稻草。那头骆驼早就被崇祯朝一块一块地卸光了骨头。

财产权的崩塌，是从"三饷"开始的。

崇祯元年1628年，为了对付东北的后金和国内的流民军，朝廷开征"辽饷"，后来又叠加"剿饷""练饷"。三项合计，每年从民间加征白银大约2000万两以上。

这在一个正常年景全国税收也不过四五百万两的农业帝国里，是什么概念？相当于帝国把民间往死里抽血。

更要命的不是税额，是征收方式。朝廷给地方官下指标，完不成就摘乌纱帽、甚至砍脑袋。地方官顶着这把刀，只好把压力转嫁给士绅地主：你家地多，你来凑。于是整个崇祯年间，士绅阶层突然发现了一个可怕的逻辑反转，家里越有钱，越是朝廷的靶子，也就越危险。

这种逻辑一旦成立，精英阶层对帝国的最后一点忠诚就开始加速蒸发。

税收带情绪，是从"按需勒索"变成"定点清算"开始的。

正常的财政税收有其内在逻辑：按田亩、按商贸流水、按固定比例，可以预期、可以规划。但崇祯年间的税，已经完全失去了"契约"属性，变成了一种应激反射，哪里冒出危机，就去哪里薅羊毛。

更绝的是，你把税乖乖交了，并不换来安全感。朝廷拿到钱后会反推：你能交这么多，说明家底还厚着，再过两个月，继续要。这形成了一个极其恶劣的"诚实税"，越顺从越暴露，越暴露越危险。

到1644年李自成攻进北京，直接对京城官绅展开"追赃助饷"，酷刑逼供、拉网清算，被拷掠出来的白银据说多达七千万两。这已经不是税收了，这是一场由流民军执行的、以公权力背书的大规模财产掠夺。

流动性死锁，富商开始埋银子隐藏资产。

崇祯年间，民间出现了一个极具时代特色的现象：大规模的银两窖藏。士绅和商人宁愿把真金白银埋进地下，封死，让它"消失"，也不愿让它以任何可见的形式存在。

这个行为的底层逻辑显而易见，白银一旦流通，就会在账目上留下痕迹，留下痕迹就意味着你是有钱人，是有钱人就等于成了下一个目标。

于是货币停止了流动，商业活动萎缩，工匠、商人的生计随之断裂，失业的底层人口以百万计地流入了流民军的队伍。流动性死锁的后果，从来不只是经济问题，它最终会变成军事问题。

三张牌全部亮出，1644年甲申年，北京城破，一个运行276年的王朝，用三个月走完了最后的程序。假如您看过《明朝那些事儿》，或许更加唏嘘不已！

第二笔账：清末，本国居民变成二等公民

如果说明末是一个帝国把刀架在自己人脖子上，那清末则更加荒诞，它演变成了外国人的财产比本国人的财产更安全的奇景。

财产权的二元化，始于甲午战败。

1895年，《马关条约》签订，清政府背上了2.3亿两白银的战争赔款，相当于当时全国财政收入的三年总和。朝廷没有钱，只能向列强借款，比如汇丰银行、德华银行、俄法银行团……借一笔，抵押一批关税或矿权；再借一笔，再抵押一笔。

与此同时，各省督抚被授权"就地筹款"，实际上拿到了几乎不受约束的地方征用权。于是可以征用的，一样一样被拿走；不可以征用的，也想方设法被拿走。

但最讽刺的事情出现了，外资企业和买办商人，因为有不平等条约的领事裁判权背书，清政府轻易不敢动。反而是毫无保护伞的本国士绅和商人，成了最好薅的羊毛。

当"做本国人"的风险远高于"挂洋人名下"，精英阶层开始疯狂地把资产托庇于洋行、英资公司、外国教会名下。这不是崇洋，这是理性的自保。这可能是有史可查的、中国历史上最早一批系统性"资产出走"的案例。

税收变成了还债机器，始于庚子赔款的天量压榨。

1901年，庚子赔款落地：4.5亿两白银，按西方国家人口折算，是"每个中国人赔一两"，分39年还清，年息4厘，本息合计9.82亿两。

清廷拼命扩充税种，"厘金"制度走向极致，货物在国内运输，每过一道关卡就被抽一层，有时候从上海到汉口一段路，要被剥七八次皮。据历史学家估算，晚清商品流通成本中，关卡厘金的摩擦损耗占到了总成本的30%到50%。

这些税不是用来修路、办学、强兵的，它就是一台吸血机器，把民间积累的财富通过层层官僚系统抽走，最终流进列强的口袋里。税收完全失去了其应有的"公共契约"属性，变成了赤裸裸的转移支付，从中国老百姓身上，转移给西方债权国。

资金外逃，走的是租界和南洋两条路。

面对国内的资产不安全感，晚清有钱人的选择出奇地一致：钱，往安全的地方搬。

上海租界的银行存款在庚子之后暴增。香港汇丰银行的华人存款在1900年到1910年间翻了数倍。南洋的华侨商业网络形成了庞大的地下汇款通道，资本在主权版图之外悄悄流转。据估计，仅1900年前后十年间，从国内流向香港、南洋的民间资本，保守估计超过数亿两白银。

更深远的后果，当商业精英和地方士绅对朝廷彻底失去信任，革命党的融资渠道反而在海外华人社区里畅通无阻。 孙中山的革命经费，很大一部分就来自南洋华侨的秘密捐款，那些钱，本来是可以留在国内流通的，是清廷的税收逻辑和财产逻辑，亲手把它们推到了革命那一边。

三张牌再次齐齐亮出，1911年，武昌城头一声枪响，清王朝的最后结算，只用了不到四个月。

历史的齿轮，从不是突然断裂的。

两场覆灭，相隔267年，但剧本惊人地相似：

都不是某一天突然就崩了。 都是先让财产权变得模糊，让有产者感到不安； 然后让税收变成带着情绪的惩罚，让顺从者感到危险； 最后让流动性完全冻结，让整个社会的经济引擎慢慢熄火。

三个信号各自缓慢地积累，某一天，它们同时到达临界点。

在那个节点之前，人人都说"还可以的，还没到那一步"。 在那个节点之后，人人都说"早就知道会这样"。

真正读懂历史的人明白：那中间的这段沉默，才是最贵的信息。

当一个时代开始让人们不敢持有、不敢流通、不敢相信明天，那些足够聪明的人已经悄悄开始做自己的准备了。

历史从来不通知你。但它一直在给你塞信号。你看没看见，是另一回事。上述交流，仅做探讨！

最后朋友提示我，海外身份和资产，早做准备！🧐

[imjustnewatai @imjustnewatai]

the wildest AI roadmap is hiding in plain sight—in OpenAI and Anthropic’s job listings.

i went through them. they aren’t merely building better chatbots.

they’re assembling 6 pieces of what could become a compounding AI flywheel:

① simulated universes

Anthropic’s “Universes” team is creating ultra-realistic training worlds where agents practice difficult, long-horizon work: handling interruptions, ambiguity, changing goals and extended interactions.

think flight simulator—but for entire jobs.

② the “video game method”

OpenAI is hiring a Synthetic RL team to train models with synthetic data, simulators and self-play.

in plain English: instead of waiting for humans to demonstrate everything, give the AI a simulated task, let it act, score the result, learn from failure and play again—potentially generating harder “levels” as it improves.

the task could be a game, coding project, research problem or computer workflow.

③ superhuman science

Anthropic says it “owns the science behind superhuman epistemics.”

epistemics simply means figuring out what is true, why it is true and which evidence supports it—not merely generating a believable answer.

another team is explicitly trying to make Claude a “superhuman life sciences research assistant” capable of literature review, database searches, bioinformatics pipelines and scientific analysis.

④ train Claude to design silicon

silicon means the physical computer chips AI runs on.

Anthropic wants RL environments where Claude learns to generate chip designs, formally verify them and optimize their physical layouts.

that matters because better AI could eventually help design better AI hardware—which provides more compute for training better AI.

⑤ personal AGI with memory

OpenAI says it wants to evolve ChatGPT from a chatbot into an “infinitely capable and personalized superassistant.”

its consumer-device researchers are working on long-term memory, evolving user models and systems that learn your preferences, goals and values over time.

not an assistant you repeatedly brief. one that gradually learns how to help you.

⑥ agents operating the AI factories

OpenAI’s Stargate role applies agents to cluster deployment, incident diagnosis, health monitoring and capacity decisions.

the listing specifically describes “closed-loop automation”:

watch the infrastructure → decide what to do → act → check the result → repeat.

connect everything and a possible loop appears:

simulated worlds
→ stronger long-horizon agents
→ superhuman research + chip design
→ better hardware and infrastructure
→ agents help operate that infrastructure
→ stronger models
→ repeat

this is what recursive self-improvement may look like in reality.

not one conscious model suddenly rewriting itself overnight—but an entire AI ecosystem improving its simulations, data, science, chips, infrastructure and training process.

important: these listings reveal research ambitions, not proof that the complete system works today or will arrive on a particular date.

but everyone expects AGI to appear as one dramatic model release.

the job boards suggest something stranger:

it may arrive as a flywheel.