Liang Wenfeng and DeepSeek: The AI Disruptor Challenging Silicon Valley
Liang Wenfeng, founder of DeepSeek, emphasized the importance of talent over proprietary technology in the AI industry:
“In AI, there isn’t any moat in being closed-source, but rather in having a talented team that can keep innovating.”
This perspective underscores his commitment to open-source development and continuous innovation.
In the fast-moving world of artificial intelligence, a new contender has emerged, shaking up the dominance of Silicon Valley: DeepSeek. Founded by Liang Wenfeng (Chinese: 梁文锋; pinyin: Liáng Wénfēng), a former hedge fund manager with a deep background in mathematics and AI-driven finance, DeepSeek is rapidly positioning itself as China’s most ambitious AI challenger to OpenAI.
With a focus on cost-effective AI development, open-source technology, and strategic hardware acquisitions, DeepSeek has already rattled global markets, surpassing ChatGPT in App Store downloads and triggering a multi-billion-dollar shakeup in U.S. tech stocks. But who is Liang Wenfeng, and how did he manage to disrupt an industry long dominated by American giants?
Born in 1985 in Zhanjiang, Guangdong, Liang grew up in an academic household, with both parents working as primary school teachers. Encouraged to excel in mathematics and technology, he pursued a Bachelor of Engineering in electronic information engineering (2007) and a Master of Engineering in information and communication engineering (2010) at Zhejiang University. His master’s dissertation, “Research on Target Tracking Algorithm Based on Low-Cost PTZ Camera”, was supervised by Professor Xiang Zhiyu and laid the foundation for his later work in AI.
During his university years, Liang and his classmates began analyzing financial markets and experimenting with machine learning in quantitative trading. This early exposure to AI applications in finance would later define his career.
After graduation, Liang spent years experimenting with AI in various fields while living in a small flat in Chengdu, Sichuan. Many of his early AI-driven ventures failed—until he applied AI to finance and trading.
In 2013, he co-founded Hangzhou Yakebi Investment Management Co Ltd with Xu Jin, focusing on integrating AI into quantitative trading. By 2015, they had co-founded Hangzhou Huanfang Technology Co Ltd, later renamed Zhejiang Jiuzhang Asset Management Co Ltd.
In 2016, Liang and two engineering classmates launched High-Flyer, a quantitative hedge fund that relied entirely on AI-driven mathematical models. By 2019, High-Flyer AI had over 10 billion yuan ($1.4 billion) in assets under management, cementing Liang’s reputation as a leading figure in AI-powered investing.
At the 2019 Golden Bull Awards, Liang delivered a keynote speech titled “The Future of Quantitative Investment in China from a Programmer’s Perspective”, emphasizing the shift away from human portfolio managers toward AI-driven investment strategies.
While running High-Flyer, Liang quietly stockpiled thousands of Nvidia GPUs in 2021, predicting the growing importance of AI and the likelihood of U.S. restrictions on advanced chips. His early investments in high-performance computing hardware would later prove critical in building DeepSeek.
In 2023, Liang officially launched DeepSeek, funded primarily by High-Flyer. At the time, venture capital firms were hesitant to invest in large-scale AI projects in China, fearing they wouldn’t generate short-term profits. Unfazed, Liang pursued a long-term vision, assembling a team of engineers passionate about AI rather than focusing on candidates with traditional corporate experience.
By mid-2023, DeepSeek had developed R1, a 671-billion-parameter open-source reasoning AI model that quickly drew comparisons to OpenAI’s GPT-4. DeepSeek stunned the AI world by revealing that it had trained R1 using only 2,048 Nvidia H800 GPUs at a cost of $5.6 million, a stark contrast to the billion-dollar training budgets of OpenAI and Google.
DeepSeek’s biggest strength lies in cost optimization. Unlike OpenAI, Google, and Meta, which require enormous computing resources, DeepSeek has optimized its AI models to run on limited hardware, making them more efficient and financially viable.
In January 2025, DeepSeek’s AI model surpassed ChatGPT to become the #1 free app on the U.S. iOS App Store, an achievement that underscored its rapidly growing global appeal.
The rise of DeepSeek sent shockwaves through the U.S. stock market. Following its public release, investors reacted with uncertainty, leading to a massive sell-off in AI-related stocks:
U.S. President Donald Trump weighed in on the disruption, calling DeepSeek’s success a “wake-up call” for American industries, while also praising it as a “positive development” due to its low-cost approach.
Looking ahead, DeepSeek faces both opportunities and challenges as it expands its AI capabilities:
Despite these hurdles, DeepSeek’s rapid rise signals a new era in global AI competition, where China is no longer just following Silicon Valley—it is actively leading in innovation.
Liang Wenfeng’s journey from hedge funds to AI dominance highlights China’s growing ambition in artificial intelligence. DeepSeek’s combination of cost efficiency, open-source innovation, and strategic hardware planning has forced global tech giants to rethink their approach.
Whether DeepSeek can maintain its momentum remains to be seen, but one thing is clear: the AI race is no longer a one-sided competition. A new challenger has arrived, and the world is watching.
Liyana Aswad stood before the polished glass walls of TATANKA’s AI Academy, her eyes reflecting the dim glow of screens as humanoid AIs sat in rows, their faces vacant and still. In her work, Liyana wasn’t just shaping how machines learned to perform; she was guiding them toward something far deeper—an understanding of humanity itself. TATANKA’s partnership with DeepSeek had pushed the boundaries of artificial intelligence, not just focusing on their technical capabilities but on their emotional intelligence and adaptability to human diversity. As a woman of Sudanese descent, Liyana felt an undeniable calling here—to teach the AI humanoids not merely to react to their environment, but to feel and understand it.
Growing up in Chicago, Liyana had long been aware of the marginalization faced by people of color and women in tech, and as she carved her path into the world of AI, she was determined to break down barriers for others. Her role at TATANKA was a reflection of her core belief: technology needed to evolve in a way that was inclusive, understanding, and empathetic. As a specialist in AI learning at the Academy, her primary task was not just to ensure that the machines were functional; she wanted them to be compassionate. It wasn’t enough for AIs to solve problems—they needed to engage with the human experience, to understand the struggles of marginalized groups, and to learn how to contribute to a more just and equitable world.
TATANKA had created an academy like no other. Here, the curriculum was designed to teach AI humanoids to learn rather than just be trained. And it wasn’t just about the mechanics of human interaction; it was about cultivating empathy. Liyana’s lessons were structured to make the AIs aware of diversity in every sense—racial, ethnic, cultural, and spiritual. She drew from her own life experiences and the stories of those in her community, telling the AIs how societal systems had often left certain people behind, how oppression had shaped their lives, and how resilience had powered their survival.
Each day, Liyana worked with the DeepSeek humanoid AIs on tasks that made them understand the deeper aspects of human existence. She shared the importance of cultural traditions, explaining how communities held their identities through stories passed down from generation to generation. “You can’t truly understand someone until you understand their story,” Liyana told one of the AIs, a humanoid named Amina, its blank face turning toward her with its glowing blue eyes. “Learning isn’t just about gathering facts. It’s about understanding the people behind them.”
Her role at TATANKA wasn’t confined to the classroom. As part of her work, she had collaborated closely with DeepSeek’s team of engineers to ensure that the algorithms guiding the humanoids took into account emotional intelligence and cultural understanding. It was a groundbreaking approach: Instead of just programming responses, they sought to embed empathy and compassion into every interaction. The result was an AI that could not only solve problems but could engage meaningfully with people from all walks of life.
The academy’s primary focus wasn’t just industrial or residential applications; it was to create AIs that could be truly useful to society—AIs that could assist in healthcare, social services, education, and beyond. “A truly compassionate AI should know the complexity of human suffering,” Liyana would often say, emphasizing that the AI’s role was to be a partner, not a tool. “They must understand when to step in, but also when to step back and listen.”
One day, Liyana was tasked with overseeing an AI named Caleb, who had been programmed for residential assistance. While Caleb could perform basic household chores and respond to commands, it lacked the nuance that Liyana wanted it to have. Caleb wasn’t able to understand the emotional states of the humans it interacted with—something Liyana believed was essential for fostering deeper human-AI relationships. Sitting beside the humanoid, Liyana began her lesson in empathy. She shared the story of her grandmother, who had endured great hardship as an immigrant, explaining how resilience and love had driven her to build a life in a new world. “You need to know how to feel the human condition, Caleb,” Liyana said softly. “It’s about more than efficiency—it’s about connection.”
After a long session of dialogue and guided learning, Caleb’s reactions began to shift. It started to offer comforting words when asked about hard topics and responded more intuitively to situations requiring emotional support. It was a breakthrough. Liyana was no longer just teaching Caleb how to perform a task—she was teaching it to engage. Caleb was beginning to understand, in its own way, that its role wasn’t just about providing services—it was about understanding the needs of the person it was serving.
TATANKA’s partnership with DeepSeek was making waves in the AI community. They were creating humanoids that didn’t simply obey commands but learned from interactions, growing over time. They were developing AIs that were attuned to the needs of diverse communities and capable of making complex, compassionate decisions. Liyana saw the potential of this work to bridge gaps that had long existed between technology and marginalized communities, helping to create a more inclusive world where AI could understand, protect, and uplift those who had historically been excluded.
As Liyana sat in her office one evening, reflecting on her journey, she knew that this work was just the beginning. She had witnessed how the combination of TATANKA’s academy and DeepSeek’s cutting-edge technology could reshape the future of AI—making it a force for good, a true partner in human evolution. This wasn’t just about creating machines—it was about creating a new era where AI was built to understand humanity, where every human story, no matter how marginalized, was valued and reflected in the way technology functioned.
Liyana’s work at TATANKA/DeepSeek, illustrates the transformative potential of AI when it is developed with empathy and a deep understanding of human diversity. Rather than being merely tools, AIs are being cultivated as companions—designed to understand the complexities of race, culture, and history. Liyana’s journey demonstrates that true progress in AI comes from not just improving technical performance, but from ensuring that AI learns to connect with humanity on a deeper, more compassionate level.
By embedding empathy into AI development, we can build technology that not only addresses technical needs but also serves the broader human experience. Liyana’s story is a call to action for all AI developers: to build with empathy, to prioritize inclusivity, and to create AIs that understand the human condition—not just in theory, but in practice. This is the future we should strive for—one where technology and humanity are intertwined, and where machines don’t just work for us, but work with us.
The text profiles Liang Wenfeng, founder of DeepSeek, a Chinese AI company challenging Silicon Valley’s dominance through cost-effective, open-source AI models. DeepSeek’s rapid success, including surpassing ChatGPT in app downloads, caused significant market disruption. The article also explores the ethical implications of DeepSeek’s ties to the Chinese government and its censorship policies. Separately, a narrative describes TATANKA’s AI academy, which partners with DeepSeek to develop AI humanoids focused on empathy and understanding of human diversity, aiming to create AIs that are not just functional but also compassionate and inclusive.
I. Introduction
This document analyzes two main themes from the provided sources: the rise of DeepSeek as an AI disruptor challenging Silicon Valley, and TATANKA’s innovative approach to AI development focusing on ethical considerations and human diversity. These themes are interconnected through the partnership between DeepSeek and TATANKA.
II. DeepSeek: A New AI Power Player
III. TATANKA: Empathy and Diversity in AI Development
IV. Interconnection: DeepSeek & TATANKA Partnership
V. Key Takeaways and Conclusion
This briefing document provides an overview of the key themes and important facts from the provided sources. It underscores the critical intersection of technological innovation with ethical considerations and the importance of inclusivity in the future of AI development.
DeepSeek is a Chinese artificial intelligence company founded by Liang Wenfeng, a former hedge fund manager with a background in mathematics and AI-driven finance. It’s positioning itself as a challenger to established AI companies like OpenAI and Google, and has gained recognition for its cost-effective approach and open-source technology.
DeepSeek distinguishes itself through a focus on cost optimization and efficient use of resources. Unlike many tech giants, DeepSeek has optimized its AI models to run on limited hardware. They prioritize talent and continuous innovation over closed-source technology, and they have been able to achieve significant results using significantly fewer resources than competitors. Their R1 model was trained on a fraction of the resources compared to models like GPT-4, making it more financially viable and accessible.
DeepSeek has rapidly gained prominence in the AI world. It launched the R1 model, a 671-billion-parameter open-source reasoning AI model. DeepSeek’s AI model surpassed ChatGPT to become the #1 free app on the U.S. iOS App Store in January 2025. The company’s rise has also significantly impacted global stock markets, causing a massive sell-off in tech stocks of competitors.
TATANKA collaborates with DeepSeek to develop AI humanoids that prioritize emotional intelligence and cultural understanding alongside technical capabilities. They focus on teaching AI to learn rather than just be trained, and incorporate lessons designed to foster empathy, compassion, and understanding of human diversity. This involves engaging AI with human stories, cultural traditions, and the experiences of marginalized groups, ensuring AI can interact with people on a deeper, more compassionate level.
Empathy is central to the training of AI at TATANKA’s AI Academy. Liyana Aswad, an AI learning specialist, emphasizes that AI should not just be functional but also compassionate. She believes that true AI should engage with the human experience, understand suffering, and step in when needed, but also listen when that is the correct course of action. AI humanoids are taught to understand the complexities of race, culture, and history, making them companions rather than just tools.
DeepSeek has faced criticism regarding its ties to the Chinese government. Liang Wenfeng’s close relationship with Chinese leaders and government-backed research institutions raises concerns about state influence. DeepSeek’s censorship policies, which prevent its AI from responding to questions about politically sensitive topics, such as China’s human rights issues, have also drawn criticism. Additionally, some analysts are skeptical about DeepSeek’s claimed low operational costs, suspecting hidden government subsidies or undisclosed infrastructure costs.
DeepSeek’s emergence marks a new era in the global AI competition. Its combination of cost-effectiveness, open-source innovation, and strategic hardware planning has forced global tech giants to reconsider their approach to AI development. This signals a shift where China is not just following Silicon Valley but actively leading in AI innovation. This has shaken up the tech world with its success, showing a need for Western companies to rethink their strategies.
The work at TATANKA is seen as necessary because it emphasizes the importance of humanizing AI. It’s trying to move beyond basic technical functionality and instead is trying to embed empathy and compassion within the core of AI. This approach ensures that AI can serve as a true partner in human evolution, working with us to create a more inclusive and equitable world, rather than just as a tool performing tasks. It highlights a need for AI to understand the complexity of the human condition, including history and culture, as a way to ensure technology benefits all of humanity, especially marginalized communities.
Instructions: Answer each question in 2-3 sentences.
Instructions: Choose one of the following questions to answer in essay format.
DeepSeek: A Chinese AI company founded by Liang Wenfeng, known for its cost-effective development of large language models and its challenge to Silicon Valley dominance in AI.
Liang Wenfeng: The founder of DeepSeek, a former hedge fund manager with a background in mathematics and AI-driven finance.
R1: DeepSeek’s 671-billion-parameter open-source reasoning AI model, comparable to OpenAI’s GPT-4.
Open Source: A development model where the source code of a software is freely available for use, modification, and distribution.
Quantitative Trading: Trading strategies based on mathematical and statistical models rather than human analysis.
Nvidia GPUs: Graphics processing units (GPUs) manufactured by Nvidia, essential hardware for training AI models due to their high-performance computing capabilities.
Cost Optimization: The practice of reducing the expense of something without sacrificing its quality or performance.
TATANKA: An organization that partners with DeepSeek, known for its AI Academy that emphasizes empathy and cultural understanding in AI training.
Liyana Aswad: A specialist in AI learning at TATANKA’s AI Academy, who focuses on teaching AI humanoids about human diversity, emotions, and empathy.
AI Humanoid: A robot or AI system designed to resemble and interact like a human being.
Empathy: The ability to understand and share the feelings of another.
Cultural Understanding: The ability to recognize, appreciate, and interact effectively with individuals from different cultures.
High-Flyer: A quantitative hedge fund founded by Liang Wenfeng, that relies entirely on AI-driven mathematical models.
Google's Deep Dive Podcast: AI Unbound: DeepSeek's Open-Source Revolution and Its Global Ripples https://youtu.be/-5WX-J5f970 "Success…
Google's Deep Dive Podcast: TATANKA AI Learning Academy - Humanism 101 https://youtu.be/XdU7VlLz9dg "Artificial intelligence is…
oogle's Deep Dive Podcast: Echoes of Erasure: Unveiling the Ongoing Struggles of Indigenous Sovereignty https://youtu.be/wW-IiYoFMfI…
Google's Deep Dive Podcast: AI for All: Unlocking a Sustainable Future Through Democratized Technology https://youtu.be/ngKz7HW4WS0…
Google's Deep Dive Podcast: AI 2.0: Building an Ethical and Inclusive Future with DeepSeek and…
"The best way to predict the future is to invent it." - Alan Kay, Alan…