Source: Grayscale; Compiled by: Deng Tong, Golden Finance
Abstract
The launch of the DeepSeek AI model has been called the "Sputnik moment," highlighting the international fight for AI hegemony and the power of open source technology.
At the same time, the news also highlights the risks associated with centralized AI development, such as data security, bias, and censorship. The risks associated with centralized AI companies can be addressed by blockchain-based AI platforms such as Bittensor.
Bittensor helps promote the development of open and global AI systems through the use of decentralized networks and economic incentives. By leveraging blockchain technology and a global network of participants, platforms like Bittensor can increase transparency, democratize access, and distribute ownership of AI systems.
Grayscale Research believes that the emergence of DeepSeek may reduce the cost and entry barriers of open source decentralized AI projects.
What Happened?
Recently, China-based startup DeepSeek launched an open-source artificial intelligence (AI) model that matches or exceeds the performance of leading models such as OpenAI’s o1. [1] Impressively, DeepSeek achieved this with significantly fewer computing resources, reportedly spending only about $5 million to train the model—a fraction of the hundreds of millions OpenAI spent training. [2] As of January 27, DeepSeek surpassed OpenAI’s ChatGPT in the Apple App Store rankings. [3]
Tech leaders are calling this AI’s “Sputnik moment”—that we may be witnessing a modern-day space race between China and the United States in the field of AI. [4] DeepSeek’s emergence led to a historic sell-off in tech stocks, wiping billions of dollars off the market value of companies such as Nvidia and Microsoft, as it forced investors to rethink their assumptions about this powerful emerging technology. [5]
However, while DeepSeek’s breakthroughs showcase the power of open source AI, they have also raised greater attention to the risks of centralized control and development of AI technology. Shortly after news of the company’s new model’s performance was announced, DeepSeek suffered a large-scale cyberattack, prompting the company to temporarily restrict user registration. [6] The incident highlighted the vulnerabilities inherent in centralized systems—including the risk that a cyberattack could disrupt service. Distributed systems can enhance network resilience by spreading responsibility across multiple entities. Decentralized development of AI models may also help reduce bias and increase transparency in this critical technology.
In this article, we will explore these risks associated with AI development and detail how decentralized AI platforms such as Bittensor are addressing them. We will also explore Bittensor’s progress to date and DeepSeek’s potential impact on the broader development of decentralized AI.
Risks of Centralized AI
Network effects and intensive capital requirements make it difficult for many AI developers outside of large tech companies (such as small companies or academic researchers) to either access the resources needed for AI development or to monetize their work. This could limit overall AI competition and innovation.
As a result, influence over this critical technology is largely concentrated in the hands of a few tech giants, raising serious questions about censorship and bias. For example, in February 2024, Google’s AI image generator Gemini was exposed for racial bias and historical inaccuracies, illustrating how companies can manipulate their models. [7] Notably, these concerns extend to DeepSeek.
This raises broader questions about AI governance. A small number of people control companies that develop a small number of models, which may increasingly shape and influence society. As the influence and importance of AI continues to grow, many worry that a single company or government could seize decision-making power over AI models that have a large impact on society, and could put up guardrails, operate behind closed doors, or manipulate the models to profit—but at the expense of the rest of society.
How can we ensure we can trust the models we use for our data? With a lack of true transparency — and with the stakes so high — how can we trust that these innovative technologies are being built in our best interests and not at our expense?
Decentralized AI and Bittensor
Enter decentralized AI: a potential solution to these challenges. By leveraging blockchain technology and a global network of participants, platforms like Bittensor can increase transparency, democratize access, and distribute ownership of AI systems.
Grayscale Research believes that decentralized AI has the potential to bring important decisions about AI development out of walled gardens and into public ownership. We believe Bittensor offers a compelling solution as a key decentralized AI platform poised to help address these risks and provide a viable alternative to centralized AI incumbents.
What is Bittensor
Bittensor is a platform that leverages decentralized networks and economic incentives to help foster the development of open and global artificial intelligence systems. It aims to create an “Internet of AI” where interconnected ecosystems are called “subnets,” each focused on a different, specific use case. Currently, Bittensor has over 50 subnets covering a wide range of applications and use cases, including video generation, AI agents, and deepfake detection. Here’s how Bittensor attempts to solve the problems associated with centralized AI:
Aligning economic incentives:Centralized AI companies prioritize shareholder value and profits, which often results in extracting value from users. In contrast, by using its TAO token, Bittensor aligns incentives between ecosystem participants, including its users and token holders.
Permissionless access to build and use AI:Many centralized AI platforms often have high barriers to entry for developers. Additionally, as AI becomes more powerful, there may be increasing restrictions on who can build or access these applications. Bittensor offers an alternative that allows permissionless access to resources to develop and use AI.
Open Source Monetization:While open source AI models like DeepSeek’s R1 and Meta’s Llama offer benefits, open source AI still faces significant difficulties in monetization and coordination. Bittensor helps solve this problem with the TAO token offering, allowing AI developers to monetize and fund their work.
We believe Bittensor’s token (TAO) is particularly compelling investment value right now for the following reasons:
Potential to solve the issues associated with centralized AI listed above.
Progress has been made to attract ecosystem investors such as Yuma and subnet builders such as Masa (data scraping and AI agent arena), Dippy (AI role-playing application), and Kaito (decentralized search).
The dynamic TAO (“dTAO”) upgrade, scheduled for February[9], will enable investing in individual subnets; we believe this could inject a new wave of liquidity into the Bittensor ecosystem.
The Broader Decentralized AI Landscape
Just recently, some may have argued that open source AI would always lag behind the best closed-source models offered by tech giants. DeepSeek demonstrates that this need not be the case in the future; critical AI innovations do not need to be done in silos or trickle down from the top.
Grayscale Research anticipates that a wide variety of decentralized AI assets could benefit. As efficiency gains are learned and applied, DeepSeek’s development could spur widespread improvements in decentralized AI. Access to DeepSeek’s high-performance open source models could reduce costs and barriers to entry for many open source decentralized AI projects, especially at the application layer. [10]
We are already seeing this happen. For example, decentralized AI agent launchpad ai16z already allows agents built using its ELIZA framework to access DeepSeek’s models. [11] On January 27, Venice.ai launched a token, a decentralized application that provides access to DeepSeek models on users’ local devices while preserving their data privacy. The token was valued at over $1 billion within two hours of its launch. [12]
Conclusion
As developments like DeepSeek continue to shape the AI landscape, international competition for technological supremacy, and society, Grayscale Research believes that we must embrace decentralized solutions that address the risks of centralization. By leveraging platforms like these, we can potentially prevent monopolistic control and build a safer future for AI.
Notes
[1] “How China’s top AI model overcame U.S. sanctions.” MIT Technology Review. January 24, 2025
[2] “How DeepSeek’s AI stacks up against OpenAI’s model.” The Wall Street Journal. January 28, 2025
[3] “China’s DeepSeek AI beats ChatGPT in the app store: Here’s what you should know.” CNBC. January 27. [4] “China’s DeepSeek AI shakes up industry and undercuts US momentum.” BBC. 28 January 2025. [5] “China’s DeepSeek AI shakes up industry and undercuts US momentum.” BBC. 28 January 2025. [6] “DeepSeek suffers ‘massive’ cyber attack after AI chatbot tops app store charts.” The Guardian. 27 January 2025. [7] “Google apologizes for ‘mistake’ after Gemini generates racially diverse Nazis.” The Verge. February 21, 2024. [8] The Independent. January 28, 2025. [9] X.com [10] “How DeepSeek is better than ChatGPT: a cost comparison.” Creole Studios. January 28, 2025. [11] “Will DeepSeek spark a major shake-up in the AI agent space? Is it time to buy the dip or get out?” AICoin.com. January 27, 2025. [12] “Venetian AI Tokens, which allow private access to DeepSeek, are valued at $1.6 billion.” TradingView. January 27, 2025. [13] Holdings subject to change without notice.
Anais
Kikyo
Weatherly
Catherine
Joy