Physics-Inspired Model Reveals How Deep Neural Networks Learn Features
Spring-block physics provides a novel perspective on how deep neural networks learn and develop features layer by layer. Devamını Oku
Yazılar
OpenAI’s GPT-5 Model Nears Release Amid High Expectations
OpenAI is on the brink of releasing GPT-5, the next-generation language model succeeding GPT-4, which powered the ChatGPT phenomenon starting in 2022. Industry insiders and early testers express cautious optimism, praising its enhanced coding and scientific problem-solving capabilities, though some say the leap from GPT-4 to GPT-5 feels less dramatic compared to the jump from GPT-3 to GPT-4.
OpenAI, backed by Microsoft and currently valued at around $300 billion, has faced challenges scaling GPT-5 due to limitations in available training data and increased complexity in training runs that can last months and are prone to hardware failures. Unlike GPT-4, which saw significant gains through increased compute power and data, GPT-5 incorporates a novel approach called “test-time compute,” directing extra processing power dynamically to solve complex reasoning and decision-making tasks.
Since the debut of ChatGPT nearly three years ago, generative AI has rapidly advanced. GPT-4 notably outperformed its predecessor by passing the simulated bar exam in the top 10%, setting a new standard in AI capabilities. Meanwhile, competitors like Google and Anthropic have developed rival models, and open-source initiatives such as Meta’s Llama 3 have narrowed the performance gap.
OpenAI CEO Sam Altman noted earlier in 2025 that GPT-5 would blend traditional large model training with test-time compute techniques, reflecting the company’s increasingly sophisticated and multifaceted AI portfolio. The broader AI industry awaits the release with anticipation, expecting GPT-5 to unlock new applications beyond conversational AI toward fully autonomous task execution.
AstraZeneca Signs $5.3 Billion AI Research Deal with China’s CSPC to Develop Chronic Disease Therapies
AstraZeneca has entered into a significant AI-driven research partnership with China’s CSPC Pharmaceutical Group worth up to $5.3 billion. The agreement aims to advance the development of therapies targeting chronic diseases, reinforcing AstraZeneca’s commitment to innovation in its second-largest market.
Under the collaboration, CSPC will conduct AI-led research in Shijiazhuang City to discover and develop pre-clinical candidates, including a small molecule oral therapy for immunological diseases. AstraZeneca will pay CSPC an initial upfront fee of $110 million and can receive up to $1.62 billion in development milestone payments, plus an additional $3.6 billion tied to sales milestones.
This partnership follows AstraZeneca’s broader efforts to strengthen its presence in China, including a $2.5 billion investment in a new R&D hub in Beijing announced earlier this year. It also builds on recent collaborations with AI-focused companies like Immunai, Qure.ai, and Tempus AI.
The deal also grants AstraZeneca rights to exclusive licenses for drug candidates identified through the joint research. This adds to a previous licensing deal signed in October, where AstraZeneca agreed to pay up to $1.92 billion to CSPC to enhance its cardiovascular drug pipeline.
CSPC, which derives about 80% of its revenue from finished drugs, is currently exploring additional licensing and collaboration opportunities, according to Morningstar analysts.
Sharon Barr, AstraZeneca executive, highlighted the collaboration as part of the company’s commitment to tackling chronic diseases affecting over two billion people worldwide.