摘要：DeepSeek, a relatively unknown Chinese AI startup, has sent shockwaves through Silicon Valley with its recent release of cutting-e

弗雷德英语笔记 跨学科深度学习

坚持每天认真做笔记！

All About DeepSeek - The Chinese AI Startup Challenging The US Big Tech

Janakiram MSV

Senior Contributor

I cover emerging technologies with a focus on infrastructure and AI

Jan 26, 2025,06:59pm EST

探秘DeepSeek：挑战美国科技巨头的中国AI初创公司

作者：贾纳基拉姆·MSV

资深撰稿人

我关注新兴技术，重点研究基础设施与人工智能领域

美国东部时间2025年1月26日下午6点59分

Lingyin Temple, Hangzhou PIXABAY

杭州灵隐寺 （图片来源：PIXABAY）

DeepSeek, a relatively unknown Chinese AI startup, has sent shockwaves through Silicon Valley with its recent release of cutting-edge AI models. Developed with remarkable efficiency and offered as open-source resources, these models challenge the dominance of established players like OpenAI, Google and Meta. This article delves into DeepSeek's innovative techniques, cost-efficient solutions and optimization strategies, analyzing their potential impact on the AI landscape and what it means for leading AI companies in the US.

DeepSeek，一家此前相对名不见经传的中国AI初创公司，凭借其近期发布的前沿AI模型，在硅谷引发了轩然大波。这些模型以极高的效率开发而成，并作为开源资源提供，对OpenAI、谷歌和Meta等行业巨头的主导地位发起了挑战。本文将深入探讨DeepSeek的创新技术、经济高效的解决方案及优化策略，分析其对AI领域可能产生的影响，以及这对美国领先的AI公司意味着什么。

DeepSeek: A Rising Star

Founded in May 2023 by Liang Wenfeng, a prominent figure in both the hedge fund and AI industries, DeepSeek operates independently but is solely funded by High-Flyer, a quantitative hedge fund also founded by Wenfeng. This unique funding model has allowed DeepSeek to pursue ambitious AI projects without the pressure of external investors, enabling them to prioritize long-term research and development. DeepSeek's team primarily comprises young, talented graduates from top Chinese universities, fostering a culture of innovation and a deep understanding of the Chinese language and culture. Notably, the company's hiring practices prioritize technical abilities over traditional work experience, resulting in a team of highly skilled individuals with a freshperspective on AI development.

DeepSeek：一颗冉冉升起的新星

DeepSeek于2023年5月由梁文峰创立，梁文峰在对冲基金和AI行业均颇具影响力。该公司虽独立运营，但唯一投资方是同样由梁文峰创立的量化对冲基金High - Flyer。这种独特的融资模式使DeepSeek能够在无外部投资者压力的情况下，开展雄心勃勃的AI项目，得以将长期研发放在首位。DeepSeek的团队主要由来自中国顶尖高校的年轻优秀毕业生组成，营造了创新的文化氛围，且团队对中文语言和文化有着深刻理解。值得注意的是，公司招聘更看重技术能力而非传统工作经验，组建了一支技术精湛、对AI开发有全新视角的团队。

DeepSeek's journey began with the release of DeepSeek Coder in November 2023, an open-source model designed for coding tasks. This was followed by DeepSeek LLM, a 67B parameter model aimed at competing with other large language models. DeepSeek-V2, launched in May 2024, gained significant attention for its strong performance and low cost, triggering a price war in the Chinese AI model market. This disruptive pricing strategy forced other major Chinese tech giants, such as ByteDance, Tencent, Baidu and Alibaba, to lower their AI model prices to remain competitive.

DeepSeek的发展始于2023年11月发布的DeepSeek Coder，这是一款专为编码任务设计的开源模型。随后推出了DeepSeek LLM，这是一个拥有670亿参数的模型，旨在与其他大型语言模型竞争。2024年5月推出的DeepSeek - V2因其出色的性能和低廉的成本备受关注，在中国AI模型市场引发了价格战。这一颠覆性的定价策略迫使字节跳动、腾讯、百度和阿里巴巴等中国其他大型科技巨头降低AI模型价格以保持竞争力。

DeepSeek-V2 was succeeded by DeepSeek-Coder-V2, a more advanced model with 236 billion parameters. It is designed for complex coding challenges and features a high context length of up to 128K tokens. This model is available through a cost-effective API, priced at $0.14 per million input tokens and $0.28 per million output tokens.

继DeepSeek - V2之后，又推出了DeepSeek - Coder - V2，这是一个更先进的模型，拥有2360亿参数。它专为解决复杂编码挑战而设计，上下文长度高达12.8万个令牌。该模型可通过高性价比的API获取，每百万输入令牌收费0.14美元，每百万输出令牌收费0.28美元。

The company's latest models, DeepSeek-V3 and DeepSeek-R1, have further solidified its position as a disruptive force. DeepSeek-V3, a 671B parameter model, boasts impressive performance on various benchmarks while requiring significantly fewer resources than its peers. DeepSeek-R1, released in January 2025, focuses on reasoning tasks and challenges OpenAI's o1 model with its advanced capabilities.

该公司的最新型号DeepSeek - V3和DeepSeek - R1进一步巩固了其颠覆者的地位。DeepSeek - V3拥有6710亿参数，在各种基准测试中表现出色，同时所需资源比同类模型显著减少。2025年1月发布的DeepSeek - R1专注于推理任务，凭借其先进的能力向OpenAI的o1模型发起挑战。

Microsoft’s Update Decision—Google Disappears On PCs, Androids, iPhones

Leak Reveals Musk’s Doge Could Be About To Trigger A Bitcoin-Inspired Price Game-Changer

UFC Parts Ways With Two Huge Fan-Favorite Fighters

更多精彩内容

- 微软的更新决策——谷歌在个人电脑、安卓和苹果手机上消失

- 泄密显示马斯克的狗狗币可能引发一场受比特币启发的价格变革

- UFC与两位深受粉丝喜爱的巨星拳手分道扬镳

DeepSeek also offers a range of distilled models, known as DeepSeek-R1-Distill, which are based on popular open-weight models like Llama and Qwen, fine-tuned on synthetic data generated by R1. These distilled models provide varying levels of performance and efficiency, catering to different computational needs and hardware configurations.

DeepSeek还提供一系列蒸馏模型，称为DeepSeek - R1 - Distill，这些模型基于Llama和Qwen等流行的开源模型，在由R1生成的合成数据上进行微调。这些蒸馏模型在性能和效率上各有不同，可满足不同的计算需求和硬件配置。

While DeepSeek has achieved remarkable success in a short period, it's important to note that the company is primarily focused on research and has no detailed plans for widespread commercialization in the near future.

尽管DeepSeek在短时间内取得了显著成功，但需要注意的是，该公司目前主要专注于研究，近期没有大规模商业化的详细计划。

DeepSeek's Strategic Partnerships

DeepSeek's success is not solely due to its internal efforts. The company has also forged strategic partnerships to enhance its technological capabilities and market reach. One notable collaboration is with AMD, a leading provider of high-performance computing solutions. DeepSeek leverages AMD Instinct GPUs and ROCM software across key stages of its model development, particularly for DeepSeek-V3. This partnership provides DeepSeek with access to cutting-edge hardware and an open software stack, optimizing performance and scalability.

DeepSeek的战略合作伙伴关系

DeepSeek的成功并非仅靠自身努力。该公司还通过建立战略合作伙伴关系来提升技术能力和市场影响力。其中一个值得关注的合作对象是AMD，一家高性能计算解决方案的领先供应商。DeepSeek在其模型开发的关键阶段，尤其是在开发DeepSeek - V3时，利用了AMD的Instinct GPU和ROCM软件。这种合作使DeepSeek能够使用最先进的硬件和开源软件栈，优化了性能和可扩展性。

DeepSeek’s Innovative Techniques

DeepSeek’s success can be attributed to several key innovations:

DeepSeek的创新技术

DeepSeek的成功得益于多项关键创新：

Reinforcement Learning

Unlike traditional methods that rely heavily on supervised fine-tuning, DeepSeek employs pure RL, allowing models to learn through trial and error and self-improve through algorithmic rewards. This approach has been particularly effective in developing DeepSeek-R1’s reasoning capabilities. In essence, DeepSeek’s models learn by interacting with their environment and receiving feedback on their actions, similar to how humans learn through experience. This allows them to develop more sophisticated reasoning abilities and adapt to new situations more effectively.

强化学习

与严重依赖监督微调的传统方法不同，DeepSeek采用纯强化学习，使模型能够通过试错学习，并通过算法奖励实现自我提升。这种方法在开发DeepSeek - R1的推理能力方面特别有效。本质上，DeepSeek的模型通过与环境交互并接收关于其行为的反馈来学习，类似于人类通过经验学习。这使它们能够发展出更复杂的推理能力，并更有效地适应新情况。

Mixture-of-Experts Architecture

DeepSeek’s models utilize an MoE architecture, activating only a small fraction of their parameters for any given task. This selective activation significantly reduces computational costs and enhances efficiency. Imagine a team of experts, each specializing in a different area. When faced with a task, only the relevant experts are called upon, ensuring efficient use of resources and expertise. DeepSeek’s MoE architecture operates similarly, activating only the necessary parameters for each task, leading to significant cost savings and improved performance.

专家混合架构

DeepSeek的模型采用专家混合（MoE）架构，对于任何给定任务，仅激活一小部分参数。这种选择性激活显著降低了计算成本并提高了效率。想象一下，有一组专家，每个专家都擅长不同领域。面对任务时，只调用相关专家，确保资源和专业知识的高效利用。DeepSeek的MoE架构运作方式类似，为每个任务仅激活必要的参数，从而大幅节省成本并提升性能。

Multi-Head Latent Attention

DeepSeek-V3 incorporates MLA, which improves the model’s ability to process data by identifying nuanced relationships and handling multiple input aspects simultaneously. Think of it as having multiple “attention heads” that can focus on different parts of the input data, allowing the model to capture a more comprehensive understanding of the information. This enhanced attention mechanism contributes to DeepSeek-V3’s impressive performance on various benchmarks.

多头潜在注意力

DeepSeek - V3融入了多头潜在注意力（MLA）技术，通过识别细微关系并同时处理多个输入方面，提高了模型处理数据的能力。可以将其想象为拥有多个 “注意力头”，能够聚焦于输入数据的不同部分，使模型能够更全面地理解信息。这种增强的注意力机制为DeepSeek - V3在各种基准测试中的出色表现做出了贡献。

Distillation

DeepSeek employs distillation techniques to transfer the knowledge and capabilities of larger models into smaller, more efficient ones. This makes powerful AI accessible to a wider range of users and devices. It’s like a teacher transferring their knowledge to a student, allowing the student to perform tasks with similar proficiency but with less experience or resources. DeepSeek’s distillation process enables smaller models to inherit the advanced reasoning and language processing capabilities of their larger counterparts, making them more versatile and accessible.

These innovative techniques, combined with DeepSeek’s focus on efficiency and open-source collaboration, have positioned the company as a disruptive force in the AI landscape.

知识蒸馏

DeepSeek采用知识蒸馏技术，将大型模型的知识和能力转移到更小、更高效的模型中。这使得强大的AI技术能够为更广泛的用户和设备所用。这就像老师将知识传授给学生，让学生能够在经验或资源较少的情况下，以类似的熟练度完成任务。DeepSeek的蒸馏过程使较小的模型能够继承较大模型的先进推理和语言处理能力，使其更具通用性和易用性。

这些创新技术，再加上DeepSeek对效率的关注和开源协作，使其在AI领域成为一股颠覆性力量。

DeepSeek’s Cost-Efficient Approach

DeepSeek’s commitment to cost-efficiency is evident in its development and pricing strategies:

DeepSeek的经济高效策略

DeepSeek在开发和定价策略上都体现了对成本效益的追求：

Reduced Training Costs

By leveraging RL and efficient architectures like MoE, DeepSeek significantly reduces the computational resources required for training, resulting in lower costs. DeepSeek-V3, for example, was trained for a fraction of the cost of comparable models from Meta. While the reported $5.5 million figure represents a portion of the total training cost, it highlights DeepSeek’s ability to achieve high performance with significantly less financial investment.

降低训练成本

通过利用强化学习和MoE等高效架构，DeepSeek显著减少了训练所需的计算资源，从而降低了成本。例如，与Meta的可比模型相比，DeepSeek - V3的训练成本只是其中的一小部分。虽然报道的550万美元仅代表总训练成本的一部分，但这凸显了DeepSeek以显著更低的资金投入实现高性能的能力。

Affordable API Pricing

DeepSeek’s API pricing is significantly lower than that of its competitors. This makes its models accessible to smaller businesses and developers who may not have the resources to invest in expensive proprietary solutions. For instance, DeepSeek-R1’s API costs just $0.55 per million input tokens and $2.19 per million output tokens, compared to OpenAI’s API, which costs $15 and $60, respectively.

实惠的API定价

DeepSeek的API定价明显低于竞争对手。这使得其模型对于可能没有资源投资昂贵专有解决方案的小型企业和开发者来说触手可及。例如，DeepSeek - R1的API每百万输入令牌收费仅0.55美元，每百万输出令牌收费2.19美元，而OpenAI的API收费分别为15美元和60美元。

Open-Source Model

DeepSeek’s open-source approach further enhances cost-efficiency by eliminating licensing fees and fostering community-driven development. This allows developers to freely access, modify and deploy DeepSeek’s models, reducing the financial barriers to entry and promoting wider adoption of advanced AI technologies.

开源模型

DeepSeek的开源方法通过消除许可费用和促进社区驱动的开发，进一步提高了成本效益。这使得开发者能够自由访问、修改和部署DeepSeek的模型，降低了进入门槛，推动了先进AI技术的更广泛应用。

DeepSeek’s Impact on the AI Landscape

DeepSeek对AI领域的影响

Competitive Impact

DeepSeek’s introduction into the AI market has created significant competitive pressure on established giants like OpenAI, Google and Meta. By offering cost-efficient and open-source models, DeepSeek compels these major players to either reduce their prices or enhance their offerings to stay relevant. This heightened competition is likely to result in more affordable and accessible AI solutions for both businesses and consumers. Additionally, DeepSeek’s disruptive pricing strategy has already sparked a price war within the Chinese AI model market, compelling other Chinese tech giants to reevaluate and adjust their pricing structures. This move underscores DeepSeek’s ability to disrupt well-established markets and influence overall pricing dynamics.

竞争影响

DeepSeek进入AI市场给OpenAI、谷歌和Meta等老牌巨头带来了巨大的竞争压力。通过提供经济高效的开源模型，DeepSeek迫使这些主要参与者要么降低价格，要么提升产品以保持竞争力。这种激烈的竞争可能会为企业和消费者带来更实惠、更易获取的AI解决方案。此外，DeepSeek的颠覆性定价策略已经在中国AI模型市场引发了价格战，促使其他中国科技巨头重新评估和调整其定价结构。此举凸显了DeepSeek扰乱成熟市场并影响整体价格动态的能力。

Impact on the AI Community

DeepSeek’s commitment to open-source models is democratizing access to advanced AI technologies, enabling a broader spectrum of users, including smaller businesses, researchers and developers, to engage with cutting-edge AI tools. This accessibility fosters increased innovation and contributes to a more diverse and vibrant AI ecosystem. By promoting collaboration and knowledge sharing, DeepSeek empowers a wider community to participate in AI development, thereby accelerating progress in the field. Moreover, DeepSeek’s open-source approach enhances transparency and accountability in AI development.

By making its models and training data publicly available, the company encourages thorough scrutiny, allowing the community to identify and address potential biases and ethical issues. This openness leads to more responsible and ethically sound AI advancements.

对人工智能社区的影响

DeepSeek对开源模型的坚持正在推动先进人工智能技术的普及，让更广泛的用户群体，包括小型企业、研究人员和开发者，都能接触到前沿的人工智能工具。这种可及性促进了创新的增加，有助于构建一个更多元、更有活力的人工智能生态系统。通过促进协作和知识共享，DeepSeek让更多的群体能够参与到人工智能开发中，从而加快该领域的发展。此外，DeepSeek的开源方法还提高了人工智能开发的透明度和可问责性。

通过公开其模型和训练数据，该公司鼓励全面审查，让社区能够识别并解决潜在的偏差和伦理问题。这种开放性带来了更负责任、更符合伦理的人工智能进步。

Shift in Focus

The success of DeepSeek highlights the growing importance of algorithmic efficiency and resource optimization in AI development. Instead of relying solely on brute-force scaling, DeepSeek demonstrates that high performance can be achieved with significantly fewer resources, challenging the traditional belief that larger models and datasets are inherently superior. This shift encourages the AI community to explore more innovative and sustainable approaches to development. Additionally, DeepSeek’s focus on efficiency has positive environmental implications. As concerns about the carbon footprint of AI continue to rise, DeepSeek’s methods contribute to more sustainable AI practices by reducing energy consumption and minimizing the use of computational resources.

关注点的转变

DeepSeek的成功凸显了在人工智能开发中算法效率和资源优化日益增长的重要性。DeepSeek并未单纯依赖强力扩展，而是证明了使用显著更少的资源也能实现高性能，这挑战了传统观念中更大的模型和数据集必然更优的看法。这种转变促使人工智能社区探索更具创新性和可持续性的开发方法。此外，DeepSeek对效率的关注还具有积极的环境意义。随着对人工智能碳足迹的担忧不断增加，DeepSeek的方法通过降低能源消耗和减少计算资源的使用，推动了更可持续的人工智能实践。

Strategic Timing

DeepSeek’s recent product launches, particularly the release of DeepSeek-R1, appear to be strategically timed to align with significant geopolitical events, such as President Trump’s inauguration. This timing suggests a deliberate effort to challenge the prevailing perception of US dominance in the AI sector and to showcase China’s burgeoning capabilities in the field. By synchronizing its releases with such events, DeepSeek aims to position itself as a formidable competitor on the global stage, highlighting the rapid advancements and strategic initiatives undertaken by Chinese AI developers.

战略时机

DeepSeek近期的产品发布，尤其是DeepSeek - R1的推出，似乎经过了精心的时机选择，与重大地缘政治事件相契合，比如特朗普总统的就职典礼。这种时机选择表明，它有意挑战美国在人工智能领域占据主导地位的普遍认知，展示中国在该领域迅速发展的能力。通过使其产品发布与这些事件同步，DeepSeek旨在将自己定位为全球舞台上的强大竞争者，彰显中国人工智能开发者的快速进步和战略举措。

Hugging Face Open R1 Replicates DeepSeek

Hugging Face has launched an ambitious open-source project called Open R1, which aims to fully replicate the DeepSeek-R1 training pipeline. This initiative seeks to construct the missing components of the R1 model’s development process, enabling researchers and developers to reproduce and build upon DeepSeek’s groundbreaking work.

The Open R1 project is designed with a clear three-step approach:

To support these efforts, the project includes comprehensive scripts for model training, evaluation, data generation and multi-stage training. By making the resources openly available, Hugging Face aims to democratize access to advanced AI model development techniques and encouraging community collaboration in AI research.

Hugging Face开源R1复现DeepSeek

Hugging Face启动了一个雄心勃勃的开源项目——Open R1，旨在完全复现DeepSeek - R1的训练流程。该计划试图构建R1模型开发过程中缺失的部分，让研究人员和开发者能够重现并在DeepSeek的开创性工作基础上继续发展。

Open R1项目采用明确的三步法：

为支持这些工作，该项目包含用于模型训练、评估、数据生成和多阶段训练的全面脚本。通过公开提供这些资源，Hugging Face旨在普及先进人工智能模型开发技术，并鼓励人工智能研究领域的社区协作。

Challenges for DeepSeek

DeepSeek面临的挑战

Compute Gap

Despite its notable achievements, DeepSeek faces a significant compute disadvantage compared to its US counterparts. This gap is further widened by US export controls on advanced chips, which restrict DeepSeek’s access to the latest hardware necessary for developing and deploying more powerful AI models. Although DeepSeek has demonstrated remarkable efficiency in its operations, having access to more advanced computational resources could accelerate its progress and enhance its competitiveness against companies with greater computational capabilities. Bridging this compute gap is essential for DeepSeek to scale its innovations and compete more effectively on a global stage.

计算能力差距

尽管DeepSeek取得了显著成就，但与美国同行相比，它在计算能力方面存在明显劣势。美国对先进芯片的出口管制进一步拉大了这一差距，限制了DeepSeek获取开发和部署更强大人工智能模型所需的最新硬件。尽管DeepSeek在运营中展现出了卓越的效率，但获得更先进的计算资源能够加快其发展进程，并提升其与计算能力更强的公司竞争的实力。缩小这一计算能力差距对于DeepSeek扩大创新规模并在全球舞台上更有效地竞争至关重要。

Market Perception

DeepSeek may encounter difficulties in establishing the same level of trust and recognition as well-established players like OpenAI and Google. Building a strong brand reputation and overcoming skepticism regarding its cost-efficient solutions are critical for DeepSeek’s long-term success. To gain wider acceptance and attract more users, DeepSeek must demonstrate a consistent track record of reliability and high performance. Enhancing its market perception through effective branding and proven results will be crucial in differentiating itself from competitors and securing a loyal customer base.

市场认知

DeepSeek在建立与OpenAI和谷歌等成熟企业同等程度的信任和认可度方面可能会遇到困难。打造强大的品牌声誉并克服对其低成本解决方案的质疑，对于DeepSeek的长期成功至关重要。为了获得更广泛的认可并吸引更多用户，DeepSeek必须展示出可靠且高性能的一贯表现。通过有效的品牌推广和实际成果提升其市场认知，对于使其在竞争中脱颖而出并赢得忠实客户群体至关重要。

Competitive Landscape

The AI market is intensely competitive, with major players continuously innovating and releasing new models. To stay ahead, DeepSeek must maintain a rapid pace of development and consistently differentiate its offerings. This requires ongoing innovation and a focus on unique capabilities that set DeepSeek apart from other companies in the field. By prioritizing the development of distinctive features and staying agile in response to market trends, DeepSeek can sustain its competitive edge and navigate the challenges of a rapidly evolving industry.

竞争格局

人工智能市场竞争异常激烈，主要参与者不断创新并推出新模型。为保持领先地位，DeepSeek必须保持快速的开发节奏，并持续使其产品具有差异化。这需要持续创新，并专注于使其区别于该领域其他公司的独特能力。通过优先开发独特功能并对市场趋势保持灵活响应，DeepSeek能够维持其竞争优势，并应对快速发展的行业带来的挑战。

Censorship

DeepSeek’s models are subject to censorship to prevent criticism of the Chinese Communist Party, which poses a significant challenge to its global adoption. In countries where freedom of expression is highly valued, this censorship can limit DeepSeek’s appeal and acceptance. Addressing these concerns is vital for DeepSeek’s international expansion. Balancing the requirements for censorship with the need to develop open and unbiased AI solutions will be crucial. Finding ways to navigate these restrictions while maintaining the integrity and functionality of its models will help DeepSeek achieve broader acceptance and success in diverse markets.

审查问题

DeepSeek的模型受到审查，以防止对中国共产党的批评，这对其在全球的推广构成重大挑战。在高度重视言论自由的国家，这种审查可能会限制DeepSeek的吸引力和接受度。解决这些问题对于DeepSeek的国际扩张至关重要。在审查要求与开发开放、无偏见的人工智能解决方案的需求之间取得平衡至关重要。找到在遵守这些限制的同时保持模型完整性和功能的方法，将有助于DeepSeek在不同市场获得更广泛的认可和成功。

Conclusion

DeepSeek's emergence as a disruptive force in the AI landscape is undeniable. Its innovative techniques, cost-efficient solutions and optimization strategies have challenged the status quo and forced established players to re-evaluate their approaches. While DeepSeek faces challenges, its commitment to open-source collaboration and efficient AI development has the potential to reshape the future of the industry. As the AI race intensifies, DeepSeek's journey will be one to watch closely.

结论

DeepSeek作为人工智能领域的颠覆力量，其崛起不可忽视。它的创新技术、低成本解决方案和优化策略挑战了现状，迫使老牌企业重新评估自身方法。尽管DeepSeek面临挑战，但其对开源协作和高效人工智能开发的坚持，有可能重塑该行业的未来。随着人工智能竞赛的加剧，DeepSeek的发展历程值得密切关注。

Follow me on Twitter or LinkedIn. Check out my website.

Janakiram MSV

Janakiram MSV (Jani) is a practicing architect, research analyst, and advisor to Silicon Valley startups.

关注我的推特或领英。查看我的网站。

贾纳基拉姆·MSV

贾纳基拉姆·MSV（贾尼）是一名执业架构师、研究分析师，也是硅谷初创企业的顾问。

文章来源：https://www.forbes.com/sites/janakirammsv/2025/01/26/all-about-deepseekthe-chinese-ai-startup-challenging-the-us-big-tech/

参考译文及笔记：英语老师弗雷德人工手打

转发请标注英文作者，中文翻译者及出处。

英语老师弗雷德践行英语教与学三十年，曾在中国，美国，加拿大多地中小学校任教：上传英语学习笔记，下载文化交际热点；左顾娱乐百相人生，右盼生活五味杂陈。只发布原创首发头条，敬请支持关注！

来源：走进科技生活