OORT: Born from Accidental Classroom Moments, Aims to Democratize AI Development

The AI sector has entered an explosive era. According to the research report "2024 AI Investment Report" by consulting firm Dealroom, global AI investment is expected to reach $65 billion, accounting for one-fifth of all venture capital.

Artificial intelligence has entered an explosive era. According to a research report "2024 AI Investment Report" by consulting firm Dealroom, global AI investment is expected to reach $65 billion, accounting for one-fifth of all venture capital. Goldman Sachs' research department also stated that global AI investment could approach $200 billion by 2025.

Thanks to the AI boom, funds are flocking to AI targets. For example, the A-share company Cambricon has surged over 560% since its low in February this year, with a market capitalization exceeding 250 billion RMB; the U.S. company Broadcom has surpassed a market value of $1 trillion, becoming the eighth largest publicly traded company in the U.S.

The combination of AI and Crypto is also showing a hot trend. During the artificial intelligence conference hosted by Nvidia, Bittensor (TAO) led with a market value of over $4.5 billion, while assets like Render (RND) and Fetch.ai (FET) have seen rapid value growth.

Following large language models, AI Agents have become the engine of this round of AI market. For instance, the token of GOAT surged over 100 times in 24 hours, and ACT rose nearly 20 times in a single day, igniting the Crypto world's enthusiasm for AI Agents.

However, behind the rapid development of AI, there are also concerns. According to an article by Dr. Max Li, founder and CEO of OORT, published in Forbes titled "AI Failures Will Surge in 2025: A Call for Decentralized Innovation," the AI industry faces numerous issues, such as data privacy, ethical compliance, and trust crises caused by centralization, which increase the risk of AI failures. Therefore, decentralized innovation has become an urgent priority.

Currently, OORT has established one of the world's largest decentralized cloud infrastructures, with network nodes covering over 100 countries, generating millions of dollars in revenue, and launching the open-source Layer 1 Olympus protocol (its consensus algorithm is "Proof of Honesty" PoH, protected by U.S. patents). Through the native token OORT, it encourages everyone to contribute data, achieving an incentive closed loop. Recently, OORT launched OORT DataHub, marking a further step towards global, diverse, and transparent data collection, laying a solid foundation for the explosion of DeAI.

OORT Born from Accidental Classroom Moments

To understand the OORT project, one must first understand the problems OORT aims to solve. This involves discussing the current bottlenecks in AI development, primarily related to data and centralization issues:

1. Disadvantages of Centralized AI

1. Lack of transparency leading to trust crises. The decision-making process of centralized AI models is often opaque, seen as "black box" operations. Users find it difficult to understand how AI systems make decisions, which can lead to severe consequences in critical applications such as medical diagnosis and financial risk control.

2. Data monopoly and unequal competition. A few large tech companies control vast amounts of data, creating a data monopoly. This makes it difficult for new entrants to obtain sufficient data to train their own AI models, hindering innovation and market competition. Additionally, data monopolies may lead to the misuse of user data, further exacerbating data privacy issues.

3. Ethical and moral risks are hard to control. The development of centralized AI has raised a series of ethical and moral issues, such as algorithmic discrimination and bias amplification. Moreover, the application of AI technology in military and surveillance fields has raised concerns about human rights, security, and social stability.

2. Data Bottleneck

1. Data desert. In the booming development of artificial intelligence, the issue of data deserts has gradually emerged as a key factor restricting further development. The demand for data from AI researchers has exploded, yet the supply of data has struggled to keep up. Over the past decade, the continuous expansion of neural networks has relied on large amounts of data for training, as seen in the development of large language models like ChatGPT. However, traditional datasets are nearing exhaustion, and data owners are beginning to restrict content usage, making data acquisition increasingly difficult.

The causes of data deserts are multifaceted. On one hand, data quality is uneven, with issues of incompleteness, inconsistency, noise, and bias severely affecting model accuracy. On the other hand, scalability challenges are significant; collecting sufficient data is costly and time-consuming, maintaining real-time data is difficult, and manual annotation of large datasets poses a bottleneck. Additionally, access and privacy restrictions cannot be ignored; data silos, regulatory constraints, and ethical issues make data collection arduous.

Data deserts have a profound impact on AI development. They limit model training and optimization, potentially forcing AI models to shift from pursuing large-scale to more specialized and efficient approaches. In industry applications, achieving precise predictions and decisions becomes challenging, hindering AI's greater role in fields like healthcare and finance.