- Theaihunter Newsletter
- Posts
- 🤖 US & Italy Cracks Down DeepSeek
🤖 US & Italy Cracks Down DeepSeek
[Inside] Nvidia RTX 5090 launch, Genomic Breakthrough & More
Dhiraj Das
February 01, 2025
Welcome Back, AI Hunters!
We hunt through hundreds of sources to deliver only the most exciting AI news and tool updates. Make sure to subscribe to our daily AI newsletter & move our email to your primary inbox.
In Today’s Edition:
🖥️ Nvidia RTX 5090, Ready to Purchase
🤖 Italy Cracks Down Chinese AI Chatbot DeepSeek
🛑 Texas First to Block DeepSeek
🔬 Genomic AI Breakthrough
️⚡️Trending AI News
NVIDIA RTX 5090
Nvidia RTX 5090, Ready to Purchase
Image Source: Future
Nvidia’s GeForce RTX 5090, launched at CES 2025 for $2,000, is the most powerful consumer GPU yet. It features 32GB of VRAM, increased CUDA cores, and improved memory bandwidth. Benchmarks show it outperforms both the RTX 4090 and AMD’s Radeon RX 7900 XTX across creative applications, making it a top choice for professionals.
Key Points:
9% faster in Premiere Pro, 17% faster in DaVinci Resolve, and 35% better in After Effects vs. RTX 4090.
Unreal Engine and Blender show up to 38% improvements over the RTX 4090.
The RTX 5090 outperforms the Radeon RX 7900 XTX, especially in ray tracing and 3D rendering.
Some software, like Redshift and V-Ray CUDA rendering, lacks full support.
$2,000 positions it as a premium option for professionals.
Insight:
Puget Systems tested the RTX 5090 across various creative applications. In Adobe Premiere Pro, the card was 9% faster than the RTX 4090. DaVinci Resolve Studio saw a 17% boost, while After Effects recorded a massive 35% improvement in 3D rendering tasks. Unreal Engine benchmarks indicated a 17% lead, and in Blender and V-Ray, the RTX 5090 was 38% faster than the 4090 and three times faster than the 3090 Ti.
However, software compatibility issues persist in Redshift and V-Ray CUDA rendering. Nvidia is expected to fix these with future driver updates.
Why It Matters?
For professionals in video editing, 3D rendering, and AI-driven applications, the RTX 5090 delivers unmatched speed and efficiency. While it comes with a premium price, it sets a new standard for high-performance GPUs, making it an essential upgrade for those needing cutting-edge power.
More On: Techradar
DEEPSEEK
Italy Cracks Down Chinese AI Chatbot DeepSeek Over Privacy Concerns
Italy's data protection authority, Garante, has ordered the Chinese AI startup DeepSeek to block its chatbot in the country due to insufficient responses to privacy concerns regarding personal data usage. The regulator raised questions about the collection, storage, and legal basis for handling data, particularly in relation to China. Despite DeepSeek’s claims, the authority's order remains in effect, with other European regulators also scrutinizing the app.
Key Points:
Garante ordered the block of DeepSeek's chatbot in Italy after concerns about privacy policy and insufficient data regarding personal data handling.
The Chinese company failed to provide satisfactory information on the data's collection, purpose, and storage.
DeepSeek claimed it was not subject to Italian regulation and had removed its app from Italian app stores.
European regulators in Ireland and France also examine the chatbot’s privacy policies.
Insight:
The Garante questioned DeepSeek about the data collected by its AI assistant, including the sources, legal basis, and storage location. The Chinese startup's response was considered insufficient, leading to the immediate blocking order.
Despite this, some Italian users who had previously downloaded the app reported that it was still functioning on their devices. DeepSeek, which has gained popularity and surpassed ChatGPT in U.S. app store rankings, argued that it should not be bound by European regulations.
Why It Matters?
This case highlights the growing scrutiny of AI technologies, particularly from non-European companies, over data protection and privacy. The European Union is intensifying efforts to regulate AI to ensure compliance with its strict data protection laws. Italy’s proactive stance could set a precedent for other European nations, reinforcing the importance of transparency and user consent in AI applications.
More On: Reuters
DEEPSEEK
Texas First to Block DeepSeek AI and Chinese Apps Over Security Concerns
Texas Governor Greg Abbott has issued a ban on the Chinese AI app DeepSeek and the social media apps Xiaohongshu (RedNote) and Lemon8 from all government-issued devices. This action positions Texas as the first state to restrict DeepSeek, citing concerns over data security and foreign influence.
Key Points:
Governor Greg Abbott's ban applies to DeepSeek, Xiaohongshu (RedNote), and Lemon8.
DeepSeek is a rising AI platform competing with OpenAI.
Xiaohongshu and Lemon8 are Chinese-owned social media apps gaining popularity in the U.S.
The ban is part of broader efforts to protect Texas’ infrastructure from foreign data harvesting.
This follows similar actions against TikTok, which remains banned on government devices in Texas and other states.
Insight:
DeepSeek, an AI startup, has made waves in the tech community, quickly gaining users in the U.S. due to its ability to compete with industry leaders like OpenAI. Xiaohongshu (RedNote) and Lemon8, both owned by ByteDance (the parent company of TikTok), have emerged as popular alternatives to TikTok, particularly following the app's temporary ban.
With a large user base, these platforms have raised concerns over their Chinese ownership and potential for data harvesting.
Why It Matters?
The ban highlights ongoing concerns about the security of personal and governmental data. By restricting access to apps owned by Chinese companies, Texas aims to safeguard its critical infrastructure from potential foreign espionage or influence. This move mirrors broader national security concerns in the U.S. regarding Chinese tech companies and their access to sensitive data through popular platforms like TikTok.
More On: The Seattle Times
CREATIVE AI MODEL
Genomic Breakthrough: AI is Redefining Chromatin Structure Prediction
MIT chemists have created ChromoGen, a generative AI model capable of predicting 3D genome structures in minutes, vastly outperforming traditional methods. By analyzing DNA sequences and chromatin accessibility data, the model enables rapid, accurate predictions that can be applied to genomic research, gene expression studies, and disease investigation.
Key Points:
ChromoGen predicts chromatin structures from DNA sequences in minutes, compared to days using traditional methods.
The model has been trained on 11 million chromatin conformations, allowing it to generate thousands of structure predictions quickly.
It provides valuable insights into chromatin structures across different cell types and their impact on gene expression.
The model shows potential for studying the effects of DNA mutations and their links to diseases.
Insight:
ChromoGen, developed by MIT chemists, integrates deep learning and generative AI to predict the three-dimensional arrangement of chromatin in cells. It analyzes DNA sequences and cell-specific chromatin accessibility data, which is key to understanding gene expression patterns.
Trained on millions of experimental structures, the model predicts multiple possible chromatin conformations for any given sequence. This rapid analysis capability allows the researchers to explore thousands of chromatin structures within minutes, making it a game-changer for genome research.
The model's predictions have been shown to closely match experimental data, even for different cell types. This suggests broad applications for exploring cell-specific chromatin variations, studying the relationship between chromatin structure and gene function, and even investigating disease-related mutations.
Why It Matters?
The ability to quickly predict chromatin structures will revolutionize how scientists study genome organization and gene expression. This technology allows for faster, more scalable genomic research, enabling the exploration of how variations in chromatin folding impact cellular functions.
Additionally, by modeling the effects of DNA mutations, researchers can gain valuable insights into diseases caused by chromatin structural changes. ChromoGen’s speed and accuracy offer new opportunities for studying the molecular mechanisms underlying various biological processes and diseases.
More On: MIT News
OPENAI
OpenAI Strikes Controversial Deal to Use AI in Nuclear Security
OpenAI has formed a partnership with the US government to use its AI models in nuclear security efforts. The initiative will give up to 15,000 scientists at US National Laboratories access to OpenAI’s AI technology, which aims to reduce the risk of nuclear war and secure nuclear materials.
Despite concerns over the flaws and potential risks of OpenAI’s AI models, the collaboration represents a significant step in the integration of AI into government security operations.
Key Points:
OpenAI partners with US National Laboratories for nuclear security.
15,000 scientists will have access to OpenAI’s AI models.
AI's focus: reducing nuclear war risk and securing nuclear materials.
Concerns over OpenAI’s AI models due to past flaws and data leaks.
This follows OpenAI’s push into government contracts, like ChatGPT Gov.
Insight:
The partnership was announced by OpenAI CEO Sam Altman at a Washington, DC event, and it involves using OpenAI’s O1 series of AI models to aid in nuclear security. These models have faced criticism for data leaks and inaccuracies, raising concerns about the AI's reliability.
Despite these issues, OpenAI continues to expand its involvement with the government, with the ChatGPT Gov platform also designed for use in government security. OpenAI is also engaged in talks for new funding rounds, potentially valuing the company at $340 billion.
Why It Matters?
This partnership marks a critical intersection of AI and national security. The use of AI to manage nuclear security comes with high stakes, given the potential consequences of errors. The integration of flawed AI models into such sensitive operations raises concerns about safety, accountability, and the AI’s ability to meet high-security standards.
As AI becomes more involved in government operations, it underscores the need for thorough evaluation and oversight, especially when it concerns something as serious as nuclear security.
More On: Futurism
Reply