DeepSeek’s Key Features

• DeepSeek R1 was developed with a budget of just $6 million, a remarkably low cost compared to billion-dollar AI projects.
• Despite restricted access to Nvidia chips, China overcame this obstacle through local manufacturing and cloud platforms.
• Released as an open-source model, DeepSeek challenges traditional proprietary AI systems.
• Secured the #1 spot on the App Store, gaining widespread user attention.

The AI Race Between China and the U.S.

The AI chip embargo imposed by the United States has pushed China to develop more innovative solutions. DeepSeek, developed despite these restrictions, challenges the U.S.’s dominance in AI. Additionally, China’s decision to relax AI regulations has paved the way for greater innovation in the sector.

NVIDIA Stocks Take a Hit!

NVIDIA shares experienced significant fluctuations due to fierce competition in the AI market and China’s advancements with next-generation AI models like DeepSeek. According to the chart, NVIDIA shares dropped from $156 on January 6th to $142 as of January 27th, reflecting a total decline of $14 (approximately 9%).

You Might Be Interested In: Elon Musk Shared, That Memecoin Flew: Up 500%!

This decline is attributed to DeepSeek’s ability to achieve high performance at low cost, reshaping the balance in the AI industry. Particularly, China’s focus on local solutions, despite its reliance on NVIDIA chips, may have raised investor concerns about NVIDIA’s future earnings.

The drop in shares highlights that China’s innovative approaches in AI technology could pose a threat to NVIDIA’s current market dominance.

DeepSeek’s Performance

According to its published research, DeepSeek delivers exceptional results, competing head-to-head with leading AI systems across various datasets:

• AIME 2024 dataset: Achieved 79.8% accuracy, demonstrating high-level success.
• Codeforces (coding challenges): Led the field with a performance of 96.3%.
• MATH-500 (mathematics): Scored an impressive 97.3%, showcasing unparalleled capability.
• MMLU (general knowledge): Delivered 90.8% accuracy, surpassing competitors.
• SWE-bench Verified (software verification): Scored 49.2%, proving its strength in challenging domains.

The launch of DeepSeek highlights China’s ability to create high-performing AI systems on a limited budget, signaling a significant shift in the global AI landscape. While the U.S. is heavily investing in initiatives like the Stargate project, which includes a $500 billion infrastructure plan, other nations such as the United Kingdom are rapidly scaling up their AI investments.

You can also freely share your thoughts and comments about the topic in the comment section. Additionally, don’t forget to follow us on our Telegram, YouTube, and Twitter channels for the latest news and updates.

The post DeepSeek: A Major Step in AI from China! appeared first on Coin Engineer.

The blueprint represents yet another attempt by the US government to control and advance artificial intelligence development. Six days ago, US lawmakers announced a bipartisan plan to support President Joe Biden’s administration by enforcing export restrictions on the best AI models developed domestically.

The bipartisan working group concluded, following months of meetings with industry experts and AI detractors, that AI investments are crucial to maintaining American competitiveness with its foreign competitors and enhancing the quality of life for Americans by enabling technology that may be able to treat certain cancers and chronic diseases.

Although it is not a formal bill or policy proposal, the roadmap lays the groundwork for future artificial intelligence legislation by offering a peek at the size and extent of what legislators and other pertinent parties see.

Enforcing “existing laws for AI,” the senators’ proposal also includes filling in any gaps or unintentional harmful bias, giving standards for testing to comprehend possible AI risks top priority, and creating use case-specific requirements for AI transparency and explainability.

In addition, the group suggested new standards for openness as AI technologies are released and research into how AI can affect employment and the American workforce.

Related: OpenAI Unveils ChatGPT’s Human-like Upgrade: GPT-4o Model

The AI Working Group is not the first to try to control the quick advancement of general artificial intelligence development and adoption as well as generative AI (genAI). To develop safety standards for artificial intelligence systems, the National Institute of Standards and Technology established the AI Safety Institute Consortium in February, bringing together more than 200 businesses.

Experts say the United States is falling behind a number of other nations, chief among them the European Union, which has led the way in regulating artificial intelligence. Pressure to catch up was applied to the United Nations in March when the EU passed a comprehensive new regulation controlling artificial intelligence across all 27 member nations.

The legislation established protections for general-purpose artificial intelligence, restricted the use of biometric identification systems by law enforcement, outlawed online social scoring, AI manipulation, or user vulnerability exploitation, and allowed users to file complaints and receive “meaningful explanations” from artificial intelligence providers.

The post US Senate Pushes $32 Billion AI Investment appeared first on Coin Engineer.