Yazılar

Baidu Set to Launch Next-Gen AI Model “Ernie 5” in 2025

/in /tarafından

Baidu, China’s leading tech company, is poised to unveil the next iteration of its AI model, Ernie 5, in the second half of 2025, according to a source familiar with the matter. The new model will introduce multimodal capabilities, enabling it to handle and convert various formats, including text, video, images, and audio.

This launch comes at a time of fierce competition in China’s AI sector, especially from the startup DeepSeek, which has gained attention for offering a reasoning model that competes with OpenAI’s GPT at a lower cost. Despite being an early adopter in AI with its Ernie model, Baidu has faced challenges in achieving widespread adoption, even though it claims that Ernie 4 rivals the capabilities of GPT-4.

Baidu’s AI models have lagged behind domestic competitors, including ByteDance’s Doubao chatbot and DeepSeek, in terms of user uptake. Baidu CEO Robin Li acknowledged at a recent Dubai conference that the rise of DeepSeek highlights the unpredictable nature of innovation. He also noted that investment in data centers and cloud infrastructure remains essential, even though DeepSeek has shown that AI models can be made more cost-efficient.

Chinese Companies Embrace DeepSeek’s AI Amid Growing Frenzy

/in /tarafından

Chinese companies, including Great Wall Motor and major telecom providers, are quickly integrating the AI model released by DeepSeek, capitalizing on its attention and breakthroughs. Great Wall Motor, China’s first listed automaker, confirmed that it had embedded DeepSeek’s AI into its connected vehicle system, branded “Coffee Intelligence.” This integration marks a significant shift as the company seeks to enhance its technological offerings.

Meanwhile, China’s Ministry of Industry and Information Technology (MIIT) announced that the country’s three largest telecom operators—China Mobile, China Unicom, and China Telecom—are collaborating with DeepSeek to promote the inclusive application of AI technology. This move is part of a larger trend as companies rush to incorporate the model into their products.

DeepSeek’s AI platform has sparked investor interest, fueling speculation about its disruptive potential across China’s tech sector. Stocks of Chinese companies tied to AI, including chipmakers, software developers, and data center operators, have surged in response to this new development. Capitalonline Data Service and MeiG Smart Technology, two listed companies, experienced significant stock price jumps after announcing their integration of DeepSeek’s AI. However, both firms have cautioned investors, stating that the impact on their future business performance remains uncertain.

Other industry giants like Tencent and Huawei have also joined the wave, revealing they have integrated DeepSeek’s model into their own offerings. The rapid adoption highlights the growing impact of DeepSeek’s AI on China’s tech landscape.

Hugging Face Works on Fully Open-Source Alternative to DeepSeek-R1 AI

/in /tarafından

Hugging Face has launched a new initiative to develop Open-R1, a fully open-source replication of the DeepSeek-R1 AI model. This move comes in response to last week’s release of DeepSeek-R1 by the Chinese AI firm DeepSeek, which made headlines for its advanced capabilities and potential to rival OpenAI’s cutting-edge models. While DeepSeek-R1 was made publicly available, it was not truly open-source, as crucial components like the training code and dataset were withheld. Hugging Face aims to bridge this gap by reconstructing these missing elements, ensuring a fully transparent and accessible alternative for the AI community.

Why Is Hugging Face Building Open-R1?

In a blog post, Hugging Face researchers outlined their motivation for replicating DeepSeek-R1. While the model’s architecture and weights were shared, key training assets were not disclosed, making it a “black-box” release. This means users can run the model locally, but they lack the necessary data and methods to recreate or modify it. By developing Open-R1, Hugging Face hopes to empower researchers and developers with a fully open framework, promoting transparency and collaborative AI advancements.

One of the critical missing pieces in DeepSeek-R1’s release is the dataset used for training, particularly in reasoning-specific tasks. Additionally, the training code that defines hyperparameters—essential for fine-tuning the model’s ability to process complex queries—remains undisclosed. Hugging Face’s initiative aims to reconstruct these elements, ensuring that developers can understand and improve upon the model rather than simply using it as a locked-down tool.

By working on Open-R1, Hugging Face is reinforcing its commitment to truly open AI development, countering the growing trend of AI models being released with limited transparency. If successful, this project could set a new standard for open-source AI, allowing researchers to study, improve, and build upon state-of-the-art models without restrictions. As AI development continues to accelerate, efforts like Open-R1 will be crucial in maintaining a balance between innovation and accessibility.