Gavin Baker Managing Partner & CIO, atreidesmgmt . 1) DeepSeek r1 is real with important nuances. Most important is the fact that r1 is so much cheaper and more efficient to inference than o1, not from the $6m training figure. r1 costs 93% less to *use* than o1 per each API, can be run locally on a high end work station and does not seem to have hit any rate limits which is wild. Simple math is that every 1b active parameters requires 1 gb of RAM in FP8, so r1 requires 37 gb of RAM. Batching massively lowers costs and more compute increases tokens/second so still advantages to inference in the cloud. Would also note that there are true geopolitical dynamics at play here and I don’t think it is a coincidence that this came out right after “Stargate.” RIP, $500 billion - we hardly even knew you. Real: 1) It is/was the #1 download in the relevant App Store category. Obviously ahead of ChatGPT; something neither Gemini nor Claude was able to accomplish. 2) It is comparable to o1 from a quality perspective although lags o3. 3) There were real algorithmic breakthroughs that led to it being dramatically more efficient both to train and inference. Training in FP8, MLA and multi-token prediction are significant. 4) It is easy to verify that the r1 training run only cost $6m. While this is literally true, it is also *deeply* misleading. 5) Even their hardware architecture is novel and I will note that they use PCI-Express for scale up. Nuance: 1) The $6m does not include “costs associated with prior research and ablation experiments on architectures, algorithms and data” per the technical paper. “Other than that Mrs. Lincoln, how was the play?” This means that it is possible to train an r1 quality model with a $6m run *if* a lab has already spent hundreds of millions of dollars on prior research and has access to much larger clusters. Deepseek obviously has way more than 2048 H800s; one of their earlier papers referenced a cluster of 10k A100s. An equivalently smart team can’t just spin up a 2000 GPU cluster and train r1 from scratch with $6m. Roughly 20% of Nvidia’s revenue goes through Singapore. 20% of Nvidia’s GPUs are probably not in Singapore despite their best efforts. 2) There was a lot of distillation - i.e. it is unlikely they could have trained this without unhindered access to GPT-4o and o1. As @altcap pointed out to me yesterday, kinda funny to restrict access to leading edge GPUs and not do anything about China’s ability to distill leading edge American models - obviously defeats the purpose of the export restrictions. Why buy the cow when you can get the milk for free?
看DS 对这个问题的解答-- can you analyze the impact of DeepSeek on Stargate The emergence of **DeepSeek**, a Chinese AI startup, has significantly disrupted the strategic goals and perceived viability of OpenAI's **$500 billion Stargate project**, which aims to secure U.S. dominance in AI through massive infrastructure investments. Below is a detailed analysis of DeepSeek's impact on Stargate: --- ### 1. **Technological Challenge to Stargate's "Scale Over Innovation" Approach** Stargate’s core premise relies on scaling computational resources (e.g., data centers, GPUs) to achieve artificial general intelligence (AGI). However, DeepSeek’s **R1 model**—built with **reinforcement learning (RL)** and **open-source frameworks**—demonstrates comparable performance to OpenAI's models at **3–5% of the cost**. - **Key Innovation**: DeepSeek’s RL-driven training reduces reliance on expensive supervised fine-tuning (SFT) and massive datasets, proving that algorithmic efficiency can offset resource-heavy methods. - **Impact**: This undermines Stargate’s foundational assumption that AGI requires "brute-force" compute scaling, raising questions about the necessity of its $500 billion budget. --- ### 2. **Financial and Market Implications** DeepSeek’s cost-effective models have rattled investor confidence in capital-intensive AI projects like Stargate: - **Stock Market Reaction**: Nvidia (a key Stargate partner) saw shares drop **14.5%** amid concerns that demand for high-end GPUs could decline if smaller, efficient models gain traction. - **Cost Comparisons**: DeepSeek’s R1 training reportedly cost **$5.58 million**, while Stargate’s initial phase requires **$100 billion**. Analysts warn that such disparities may pressure AI firms to justify exorbitant spending. - **Open-Source Disruption**: By open-sourcing its models, DeepSeek erodes the subscription-based revenue moats of companies like OpenAI, forcing them to seek alternative advantages (e.g., infrastructure control via Stargate). --- ### 3. **Strategic Shift for OpenAI and Stargate** Stargate is partly a defensive move to counter threats like DeepSeek: - **Moat Building**: OpenAI’s ChatGPT once held a competitive edge, but DeepSeek’s open-source models have narrowed the gap. Stargate aims to create a new moat through infrastructure monopolization. - **Focus on Compute**: Stargate’s emphasis on data centers and energy supply reflects OpenAI’s belief that AGI requires unmatched compute access. However, critics argue this overlooks algorithmic breakthroughs like DeepSeek’s. - **Geopolitical Signaling**: The project’s U.S.-centric focus contrasts with DeepSeek’s rise, which highlights China’s ability to innovate under resource constraints (e.g., U.S. chip export bans). --- ### 4. **Industry and Geopolitical Repercussions** - **AI Democratization**: DeepSeek’s open-source strategy lowers entry barriers for startups, challenging Stargate’s goal of consolidating power among U.S. tech giants. - **China’s AI Ascent**: DeepSeek’s success exemplifies China’s shift from imitation to innovation, pressuring the U.S. to accelerate Stargate’s deployment to maintain dominance. - **Policy Risks**: Analysts suggest U.S. restrictions on AI tech exports could backfire by incentivizing Chinese firms to innovate faster—a dynamic Stargate must navigate. --- ### 5. **Long-Term Industry Shifts** DeepSeek’s rise signals broader trends that could reshape AI development: - **Efficiency Over Scale**: The industry is pivoting toward techniques like **mixture-of-experts (MoE)** and distillation to reduce costs, challenging Stargate’s capital-intensive model. - **Sustainability Concerns**: DeepSeek’s low-energy methods align with growing scrutiny of AI’s environmental impact, whereas Stargate’s energy demands may face criticism. - **Market Competition**: Price wars triggered by DeepSeek in China (e.g., ByteDance and Tencent slashing AI model prices) may spread globally, pressuring Stargate’s ROI projections. --- ### Conclusion DeepSeek’s impact on Stargate is multifaceted: it exposes vulnerabilities in resource-heavy AI strategies, accelerates global competition, and forces a reevaluation of what drives technological leadership. While Stargate aims to secure U.S. dominance through infrastructure, DeepSeek’s success underscores that **innovation, cost efficiency, and open collaboration** may ultimately define the future of AI. For OpenAI, balancing Stargate’s scale with adaptive innovation will be critical to avoid being outpaced by leaner, agile competitors.The emergence of **DeepSeek**, a Chinese AI startup, has significantly disrupted the strategic goals and perceived viability of OpenAI's **$500 billion Stargate project**, which aims to secure U.S. dominance in AI through massive infrastructure investments. Below is a detailed analysis of DeepSeek's impact on Stargate: --- ### 1. **Technological Challenge to Stargate's "Scale Over Innovation" Approach** Stargate’s core premise relies on scaling computational resources (e.g., data centers, GPUs) to achieve artificial general intelligence (AGI). However, DeepSeek’s **R1 model**—built with **reinforcement learning (RL)** and **open-source frameworks**—demonstrates comparable performance to OpenAI's models at **3–5% of the cost**. - **Key Innovation**: DeepSeek’s RL-driven training reduces reliance on expensive supervised fine-tuning (SFT) and massive datasets, proving that algorithmic efficiency can offset resource-heavy methods. - **Impact**: This undermines Stargate’s foundational assumption that AGI requires "brute-force" compute scaling, raising questions about the necessity of its $500 billion budget. --- ### 2. **Financial and Market Implications** DeepSeek’s cost-effective models have rattled investor confidence in capital-intensive AI projects like Stargate: - **Stock Market Reaction**: Nvidia (a key Stargate partner) saw shares drop **14.5%** amid concerns that demand for high-end GPUs could decline if smaller, efficient models gain traction. - **Cost Comparisons**: DeepSeek’s R1 training reportedly cost **$5.58 million**, while Stargate’s initial phase requires **$100 billion**. Analysts warn that such disparities may pressure AI firms to justify exorbitant spending. - **Open-Source Disruption**: By open-sourcing its models, DeepSeek erodes the subscription-based revenue moats of companies like OpenAI, forcing them to seek alternative advantages (e.g., infrastructure control via Stargate). --- ### 3. **Strategic Shift for OpenAI and Stargate** Stargate is partly a defensive move to counter threats like DeepSeek: - **Moat Building**: OpenAI’s ChatGPT once held a competitive edge, but DeepSeek’s open-source models have narrowed the gap. Stargate aims to create a new moat through infrastructure monopolization. - **Focus on Compute**: Stargate’s emphasis on data centers and energy supply reflects OpenAI’s belief that AGI requires unmatched compute access. However, critics argue this overlooks algorithmic breakthroughs like DeepSeek’s. - **Geopolitical Signaling**: The project’s U.S.-centric focus contrasts with DeepSeek’s rise, which highlights China’s ability to innovate under resource constraints (e.g., U.S. chip export bans). --- ### 4. **Industry and Geopolitical Repercussions** - **AI Democratization**: DeepSeek’s open-source strategy lowers entry barriers for startups, challenging Stargate’s goal of consolidating power among U.S. tech giants. - **China’s AI Ascent**: DeepSeek’s success exemplifies China’s shift from imitation to innovation, pressuring the U.S. to accelerate Stargate’s deployment to maintain dominance. - **Policy Risks**: Analysts suggest U.S. restrictions on AI tech exports could backfire by incentivizing Chinese firms to innovate faster—a dynamic Stargate must navigate. --- ### 5. **Long-Term Industry Shifts** DeepSeek’s rise signals broader trends that could reshape AI development: - **Efficiency Over Scale**: The industry is pivoting toward techniques like **mixture-of-experts (MoE)** and distillation to reduce costs, challenging Stargate’s capital-intensive model. - **Sustainability Concerns**: DeepSeek’s low-energy methods align with growing scrutiny of AI’s environmental impact, whereas Stargate’s energy demands may face criticism. - **Market Competition**: Price wars triggered by DeepSeek in China (e.g., ByteDance and Tencent slashing AI model prices) may spread globally, pressuring Stargate’s ROI projections. --- ### Conclusion DeepSeek’s impact on Stargate is multifaceted: it exposes vulnerabilities in resource-heavy AI strategies, accelerates global competition, and forces a reevaluation of what drives technological leadership. While Stargate aims to secure U.S. dominance through infrastructure, DeepSeek’s success underscores that **innovation, cost efficiency, and open collaboration** may ultimately define the future of AI. For OpenAI, balancing Stargate’s scale with adaptive innovation will be critical to avoid being outpaced by leaner, agile competitors.
重要的是，只有7B！普通电脑都能运行试试了
不会的，烙印善于溜须拍马抢别人功劳，在公司比埋头苦干的老中混得好 最后要裁人都是老中优先
反智教育这么多年，每天不是debate, 就是leadership, 不是说这些不重要，但是整天搞这个，高中生连个位数乘除法，不按计算器都不知道答案。学校里的气氛就是崇拜体育好的，然后嘲笑数学好的同学。
现在终于收到反噬了。
十年育树，百年育人。这波美国不知道要多久才才能跟上来。
美帝大学招生如果继续深陷各种非academic EC不能自拔的话只会越来越拉垮
Real: 1) It is/was the #1 download in the relevant App Store category. Obviously ahead of ChatGPT; something neither Gemini nor Claude was able to accomplish. 2) It is comparable to o1 from a quality perspective although lags o3. 3) There were real algorithmic breakthroughs that led to it being dramatically more efficient both to train and inference. Training in FP8, MLA and multi-token prediction are significant. 4) It is easy to verify that the r1 training run only cost $6m. While this is literally true, it is also *deeply* misleading. 5) Even their hardware architecture is novel and I will note that they use PCI-Express for scale up.
Nuance: 1) The $6m does not include “costs associated with prior research and ablation experiments on architectures, algorithms and data” per the technical paper. “Other than that Mrs. Lincoln, how was the play?” This means that it is possible to train an r1 quality model with a $6m run *if* a lab has already spent hundreds of millions of dollars on prior research and has access to much larger clusters. Deepseek obviously has way more than 2048 H800s; one of their earlier papers referenced a cluster of 10k A100s. An equivalently smart team can’t just spin up a 2000 GPU cluster and train r1 from scratch with $6m. Roughly 20% of Nvidia’s revenue goes through Singapore. 20% of Nvidia’s GPUs are probably not in Singapore despite their best efforts. 2) There was a lot of distillation - i.e. it is unlikely they could have trained this without unhindered access to GPT-4o and o1. As @altcap pointed out to me yesterday, kinda funny to restrict access to leading edge GPUs and not do anything about China’s ability to distill leading edge American models - obviously defeats the purpose of the export restrictions. Why buy the cow when you can get the milk for free?
从今天起开始学高考，成绩第一，理工科优先，什么破体育领导船统统废弃 二十年能赶上中国
Chinese spy这个词都快成笑话了吧。
无比的感谢小红书这一波。太是时候了。
小红书现在简直就是全世界人民大团结。
就这么寸，一堆（就连华人上有些id都看不上，攻击人家底层）的美国底层老百姓忽然知道了真相，觉悟了。
而且闹到了全世界都知道。
这种动不动就chinese spy的把戏耍不动了。
虽然没有对外公布，但他们的老本行就是做量化对冲基金，搞AI只是副业
我其实觉得业界的大牛未必不知道，跟搞基建投资越高油水越大一样，只有把泡泡吹大才能多捞钱，顺便吓退小的竞争对手。 Deep seek 打破了这种信息差，对AI长期发展肯定是好的。
裁完公司直接关门 老板在这种紧要关头应该也不会那么sb了
The emergence of **DeepSeek**, a Chinese AI startup, has significantly disrupted the strategic goals and perceived viability of OpenAI's **$500 billion Stargate project**, which aims to secure U.S. dominance in AI through massive infrastructure investments. Below is a detailed analysis of DeepSeek's impact on Stargate:
---
### 1. **Technological Challenge to Stargate's "Scale Over Innovation" Approach** Stargate’s core premise relies on scaling computational resources (e.g., data centers, GPUs) to achieve artificial general intelligence (AGI). However, DeepSeek’s **R1 model**—built with **reinforcement learning (RL)** and **open-source frameworks**—demonstrates comparable performance to OpenAI's models at **3–5% of the cost**. - **Key Innovation**: DeepSeek’s RL-driven training reduces reliance on expensive supervised fine-tuning (SFT) and massive datasets, proving that algorithmic efficiency can offset resource-heavy methods. - **Impact**: This undermines Stargate’s foundational assumption that AGI requires "brute-force" compute scaling, raising questions about the necessity of its $500 billion budget.
---
### 2. **Financial and Market Implications** DeepSeek’s cost-effective models have rattled investor confidence in capital-intensive AI projects like Stargate: - **Stock Market Reaction**: Nvidia (a key Stargate partner) saw shares drop **14.5%** amid concerns that demand for high-end GPUs could decline if smaller, efficient models gain traction. - **Cost Comparisons**: DeepSeek’s R1 training reportedly cost **$5.58 million**, while Stargate’s initial phase requires **$100 billion**. Analysts warn that such disparities may pressure AI firms to justify exorbitant spending. - **Open-Source Disruption**: By open-sourcing its models, DeepSeek erodes the subscription-based revenue moats of companies like OpenAI, forcing them to seek alternative advantages (e.g., infrastructure control via Stargate).
---
### 3. **Strategic Shift for OpenAI and Stargate** Stargate is partly a defensive move to counter threats like DeepSeek: - **Moat Building**: OpenAI’s ChatGPT once held a competitive edge, but DeepSeek’s open-source models have narrowed the gap. Stargate aims to create a new moat through infrastructure monopolization. - **Focus on Compute**: Stargate’s emphasis on data centers and energy supply reflects OpenAI’s belief that AGI requires unmatched compute access. However, critics argue this overlooks algorithmic breakthroughs like DeepSeek’s. - **Geopolitical Signaling**: The project’s U.S.-centric focus contrasts with DeepSeek’s rise, which highlights China’s ability to innovate under resource constraints (e.g., U.S. chip export bans).
---
### 4. **Industry and Geopolitical Repercussions** - **AI Democratization**: DeepSeek’s open-source strategy lowers entry barriers for startups, challenging Stargate’s goal of consolidating power among U.S. tech giants. - **China’s AI Ascent**: DeepSeek’s success exemplifies China’s shift from imitation to innovation, pressuring the U.S. to accelerate Stargate’s deployment to maintain dominance. - **Policy Risks**: Analysts suggest U.S. restrictions on AI tech exports could backfire by incentivizing Chinese firms to innovate faster—a dynamic Stargate must navigate.
---
### 5. **Long-Term Industry Shifts** DeepSeek’s rise signals broader trends that could reshape AI development: - **Efficiency Over Scale**: The industry is pivoting toward techniques like **mixture-of-experts (MoE)** and distillation to reduce costs, challenging Stargate’s capital-intensive model. - **Sustainability Concerns**: DeepSeek’s low-energy methods align with growing scrutiny of AI’s environmental impact, whereas Stargate’s energy demands may face criticism. - **Market Competition**: Price wars triggered by DeepSeek in China (e.g., ByteDance and Tencent slashing AI model prices) may spread globally, pressuring Stargate’s ROI projections.
---
### Conclusion DeepSeek’s impact on Stargate is multifaceted: it exposes vulnerabilities in resource-heavy AI strategies, accelerates global competition, and forces a reevaluation of what drives technological leadership. While Stargate aims to secure U.S. dominance through infrastructure, DeepSeek’s success underscores that **innovation, cost efficiency, and open collaboration** may ultimately define the future of AI. For OpenAI, balancing Stargate’s scale with adaptive innovation will be critical to avoid being outpaced by leaner, agile competitors.The emergence of **DeepSeek**, a Chinese AI startup, has significantly disrupted the strategic goals and perceived viability of OpenAI's **$500 billion Stargate project**, which aims to secure U.S. dominance in AI through massive infrastructure investments. Below is a detailed analysis of DeepSeek's impact on Stargate:
---
### 1. **Technological Challenge to Stargate's "Scale Over Innovation" Approach** Stargate’s core premise relies on scaling computational resources (e.g., data centers, GPUs) to achieve artificial general intelligence (AGI). However, DeepSeek’s **R1 model**—built with **reinforcement learning (RL)** and **open-source frameworks**—demonstrates comparable performance to OpenAI's models at **3–5% of the cost**. - **Key Innovation**: DeepSeek’s RL-driven training reduces reliance on expensive supervised fine-tuning (SFT) and massive datasets, proving that algorithmic efficiency can offset resource-heavy methods. - **Impact**: This undermines Stargate’s foundational assumption that AGI requires "brute-force" compute scaling, raising questions about the necessity of its $500 billion budget.
---
### 2. **Financial and Market Implications** DeepSeek’s cost-effective models have rattled investor confidence in capital-intensive AI projects like Stargate: - **Stock Market Reaction**: Nvidia (a key Stargate partner) saw shares drop **14.5%** amid concerns that demand for high-end GPUs could decline if smaller, efficient models gain traction. - **Cost Comparisons**: DeepSeek’s R1 training reportedly cost **$5.58 million**, while Stargate’s initial phase requires **$100 billion**. Analysts warn that such disparities may pressure AI firms to justify exorbitant spending. - **Open-Source Disruption**: By open-sourcing its models, DeepSeek erodes the subscription-based revenue moats of companies like OpenAI, forcing them to seek alternative advantages (e.g., infrastructure control via Stargate).
---
### 3. **Strategic Shift for OpenAI and Stargate** Stargate is partly a defensive move to counter threats like DeepSeek: - **Moat Building**: OpenAI’s ChatGPT once held a competitive edge, but DeepSeek’s open-source models have narrowed the gap. Stargate aims to create a new moat through infrastructure monopolization. - **Focus on Compute**: Stargate’s emphasis on data centers and energy supply reflects OpenAI’s belief that AGI requires unmatched compute access. However, critics argue this overlooks algorithmic breakthroughs like DeepSeek’s. - **Geopolitical Signaling**: The project’s U.S.-centric focus contrasts with DeepSeek’s rise, which highlights China’s ability to innovate under resource constraints (e.g., U.S. chip export bans).
---
### 4. **Industry and Geopolitical Repercussions** - **AI Democratization**: DeepSeek’s open-source strategy lowers entry barriers for startups, challenging Stargate’s goal of consolidating power among U.S. tech giants. - **China’s AI Ascent**: DeepSeek’s success exemplifies China’s shift from imitation to innovation, pressuring the U.S. to accelerate Stargate’s deployment to maintain dominance. - **Policy Risks**: Analysts suggest U.S. restrictions on AI tech exports could backfire by incentivizing Chinese firms to innovate faster—a dynamic Stargate must navigate.
---
### 5. **Long-Term Industry Shifts** DeepSeek’s rise signals broader trends that could reshape AI development: - **Efficiency Over Scale**: The industry is pivoting toward techniques like **mixture-of-experts (MoE)** and distillation to reduce costs, challenging Stargate’s capital-intensive model. - **Sustainability Concerns**: DeepSeek’s low-energy methods align with growing scrutiny of AI’s environmental impact, whereas Stargate’s energy demands may face criticism. - **Market Competition**: Price wars triggered by DeepSeek in China (e.g., ByteDance and Tencent slashing AI model prices) may spread globally, pressuring Stargate’s ROI projections.
---
### Conclusion DeepSeek’s impact on Stargate is multifaceted: it exposes vulnerabilities in resource-heavy AI strategies, accelerates global competition, and forces a reevaluation of what drives technological leadership. While Stargate aims to secure U.S. dominance through infrastructure, DeepSeek’s success underscores that **innovation, cost efficiency, and open collaboration** may ultimately define the future of AI. For OpenAI, balancing Stargate’s scale with adaptive innovation will be critical to avoid being outpaced by leaner, agile competitors.
是杭州的公司，浙大的吧？
应该是大部分用Nvidia的芯片，在meta的llama基础上开始的，不算from scratch，成本上省了一大块。以后会不会更多用到华为的就不晓得了。
+1，deepseek这次让很多小公司看到了搞ai的可能性，搞的大家蠢蠢欲动，长期是利好卖铲子的。老黄这么敏锐不可能错过这个机会。
难道Apple又堵对了？
薄利多销，钱没那么好赚了
喜大普奔, 讨厌Sam
主力都是清华北大的
从前都是会来美国读博然后去湾区的
感谢贸易战川建国拜振华把这些娃都留给中国了
现在大家都知道用不到那么多英伟达超级贵的显卡了，都要跟中国拼成本低的算力，并且DS已经证明了算力可以做的很便宜，英伟达想再保持现在这么大的市值估计是难了。
我不敢说具体到这些娃是不是如此，但我相信类似的故事发生在一些娃身上
两国赶紧休战吧，一个需要年轻的高品质大脑，一个需要对方的高品质教育
说起来中国这个国运也是没法说了，半导体封锁和AI技术封锁两道大闸看似牢牢的，结果一帮小孩子的side project硬给你同时把两把锁给冲了个稀里哗啦
政客想主导科技和技术的走向，可谁知道这边的封锁、那边回应封锁的大基金，都不是主角。
台积电目前超越了历史单日最大跌幅，已经来到跌15%
显然不能啊。 这个是inside trading了
现在就算直接给他们绿卡，人家可能也不稀罕了。这帮年轻人在国内更有归属感，赚的也多，生活方便安全，美国的零元购，政治正确，大麻合法化，白人优先化，人家凭什么过来听一帮只会做PPT的瞎忽悠
当年就是啊，TOP2简直就是留美预备学校。我们班一半儿在美国
?! “两国赶紧休战吧，一个需要年轻的高品质大脑，一个需要对方的高品质教育”
美国没有高品质大脑，也没有高品质教育。这些中国都完爆美国了。
人家几十个孩子，做量化业务，前几年管理的资产都上百亿了， 人家来美国干啥啊？ 哪个不早财富自由了啊！
美帝高品质大脑都去搞医法商这些虚头巴脑来钱快的行业了
Apple设备用的chatgpt.这波openai肯定要降价，省钱了啊
不算吧？他们又没有英伟达的内部消息？
不在于买的谁的股票，而在于买股票利用了公众无法获得的信息。如果是这样去做空了英伟达，SEC应该是可以制裁的吧。至于制裁是否实际有效就不知道了。
是啊，美国现在还有啥，自由民主吗🙂↕️
创始人梁文峰是浙大毕业的。
好奇这个图怎么做的
只要是自己研究出来的非公开信息，而不是从公司内部打听到的，就可以用
不违法
过两年这些玩家腰斩是不是哪个洪波州检察长要跳出来向中国索赔2500亿😂
去年到处是对中国电车产能过剩批判的时候我就说了： 几年内会对AI同等批判， 没想到我还是保守了， 这么快
真的拉夸 光leadership 没人学基础理工 领导谁啊 领导了也没有人落实
这波真的叫我想起一句陈年cliche,机会总是留给有准备的人。。。
中国几千年封建统治基本靠的就是各种思想教育，搞意识形态，文科治国。
近代被西方闯进家门，踢翻在地，狠狠抽了一顿后，彻底觉悟了，从此走全民数理化道路。从建国教育制度大改革开始，到现在走了70多年。全社会数理化人才的储备那是无恙无恙。
然后工业技术革命早了100年，靠科技发家的西方国家开始自废武功，搞意识心形态是至高无上这一套了。要不是依然有源源不断的高科技人才移民，就凭美国自己的教育体系，连合格的工程师都要凑不齐了吧了。
昨天看了Ronny chieng的脱口秀，美国人各个高喊愿意die for the country,却不愿意do math homework for the country.
现实真讽刺。
风水轮流转， 真的是感叹 欧美再这么意识形态至上下去， 今后俩百年要逐步清算白人前300年给世界各地人民带来的苦难了
我感觉对美国的教育滤镜可能以后也要碎了……
怎么会是inside trading？ 又不是做空自家股票
以前学美国搞什么素质教育快乐教育， 都搞过了效果一坨屎
到最后中国美国科技前沿依然是小镇做题家们
这个Outside，当然不是inside trading，合法合规靠实力赢的
哪有70多年，改开以后40多年吧，改开以前都在搞政治斗争
领导船可以发射到太空啊
金主爸爸让他们吃屎
这么看来， Alex wang 前几天发疯， 还算是有“vision” ， 估计已经被金主喂过屎了， 气不打一出来发疯
医法都不是来钱快的行业
😂
前面几十年， 大学生还得上山下乡 接受贫下中农再教育呢。
你可以试下问GPT以色列在加沙屠杀多少妇女儿童
这群孩子在中国搞科技创新，基本情商还是有的
你想了解敏感话题出门右拐大纪元，各种奇葩故事保准满足你心理需求
人家源代码都给了，自己把限制删一下就是了。
就连美国新闻都说是half baked稍意思下的 限制。
那华尔街多少公众无法获得的消息，佩洛西多少公众无法获得的消息😂
你和美国的Woke AI 比？
除掉敏感话题的内容应该超过80%吧。
这不是所有大公司的共性吗？我记得FB、苹果、谷歌刚开始都搞过，乔布斯被苹果踢出去两次，对外当然说得冠冕堂皇，不适合管理云云。
美国这个样子，中国这个样子，学子们傻啊？
但是昨晚我刚看到黄卖了很多股票的报道，我不做股票，就偶然看到。
你就别在这里撒谎了，ChatGPT你搜一下Gaza的问题，答案和信源清晰的很。你真是坏透了
搞政治斗争主要是社会，
学校教育还是一直都是数理化为主，虽然不可避免受到影响（数学书前面先来一段毛主席语录这种），尤其是文革那几年，但是只要上课，还是数理化占比重很大。。包括大学院校的建立，系别和课程设置。
你比一下美国和中国基础教育课程安排就可以看出来了。
美国文科从小学起，阅读，写作，拼读，grammar,都是分开的课时，分开考试，合起来比数学课时多的多。
中国所有这些都是语文一门课，和数学基本一比一的课时。
到初高中就更不用说了。中国数理化生物是不同的课程，各有各的课时，数学分代数和几何是两门，美国代数几何都是数学一门课，占一个课时，科学一门包括物理化学生物。共用一个课时。
更别提大学录取里各种spike,leadership要占据多少课外时间了。
时间分配在哪里，效果就在哪里。
难说，昨天我看到老黄卖了很多股票的新闻。佩洛西估计也是。
支持，股票跌了，看看码公是不是还矫情10M都退不了休，呵呵
不是国内都不能上google吗，想知道他们train AI的资源哪里来的
嗯，后来test了几个不涉及政治的科学话题，回答的很好
不要嫉妒哈。博通英伟达股票已在高点震荡数月，估计马工硅工早就换成湾区房产，还是赢麻了😂
总统全家发币捞钱，科技精英跪拜权力忽悠牛马捞钱，牛马洗脑洗得直叫幸福感恩。
同感。这样强大的文化自信，这样一骑绝尘的硬核创新，少年强则国强，这个古老的国家，经过了激荡三十年，到底养出了迥然不同的年轻一代，带领这个国家走向可能的新生。上一次黑悟空我就有骄傲之感。同祝愿，祖国好。
是啊，我也觉得这两条美国都没有啊🙂↔️
144 - 现在122.。。。 你想什么呢？
而且今天做空的已经跑了。
我也是为这些小小少年感到骄傲，有智商有情怀，少年强则国强
那是老百姓上不了，要是工作需要都是行的
同感啊！中国这代年轻人真不错！能力强，聪明，肯吃苦，勤奋，真不错！越看越好！真是有种老姨妈看外甥的感觉！越看越觉得好。
你牛！一天损失几十万，还心情特别好。你就是那种为国家抛头颅洒热血的烈士的性格儿吧？
英伟达这个公司长期差不了 但是股价嘛，是另外一个维度的事，估值把今后10年可能挣得钱都乐观地price in 了， 泡泡爆了而已
确实，国内很多年来的气氛就是重理轻文，厉害的娃绝大部分都去理工科了。我们学校普通C9，当年毕业的时候理工科找工作比较容易，很多现在看起来很高大上的单位要求就是能毕业。文科生找工作的行情就差多了。
本来就是搞个旗号圈钱，结果没想到人民大众热情如此高涨，成为时尚文化必备单词
学好数理化，走遍天下都不怕！真理
句句在理，现在身边中国人家长都是：我们不推学术，推leadership 推情商，和一群老美去抢那些搞social-emotional learning 的老牌贵族私立学校，～ 而且还学老美说不要把娃养成nerd
都是被洗脑了，美国的culture 需要被重新定义了
对啊，我回去到处都是VPN的广告，只是大家翻墙一通发现。。除了黄片没啥好折腾的。
新升级的那Iphone软件是什么东东啊！ 纯粹脱裤子放屁， 搞得跟迷宫一样
别忘了， 那些综合性理工科大学和各个专业都是谁建的， 那些教授都是什么时候培养出来的。 你说这话就好像你吃了第三个馒头才饱了， 以前的两个都不算数一样
好歹看一下新闻，你问的都有答案，女大为啥跌是今天大家都在谈的主题。另外，女大造cpu的？！