Research Finds GPT and Other AI Models Cannot Analyze SEC Filings

Translated data: A startup company's research has found that GPT and other AI models often fail to correctly answer questions when analyzing SEC filings. AI models must demonstrate higher performance in regulated industries such as finance to be practically applicable. Additionally, the non-determinism and uncertainty of AI models are among the challenges for their application.

AI Email Service Provider Superhuman Announces Acquisition of AI Detection Startup GPTZero

AI email provider Superhuman has acquired GPTZero, an AI-detection startup founded by Princeton grad Edward Tian and CTO Alex Cui. The acquisition integrates detection tech to identify and flag AI-generated low-quality text while helping users refine writing to sound more human. Financial terms were not disclosed.....

Report: DeepSeek Completes A-Round Financing of 51 Billion Yuan, with Giants Like Tencent and JD.com Participating

DeepSeek completed an A-round financing of approximately 51 billion yuan, with its valuation surging to 400 billion yuan. Due to the highly promising market prospects, the financing competition was intense, and it has shifted from seeking investors to companies screening participation qualifications, with a top-tier investment team ultimately investing.

Gu Quanquan Confirms Departure from ByteDance's Seed Team, Previously Led the Development of SeedFold and Seed2.0 Training System

Gu Quanquan, a core researcher at ByteDance's Seed team, has confirmed her departure. She shared her research achievements in AI drug discovery and pre-training of large language models over the past three years on a social platform. The bio-molecular structure prediction model SeedFold, which she led the development of, performed excellently in multiple public benchmark tests. This departure comes as ByteDance's AI business accelerates its commercialization, drawing attention to the emerging trend of AI for Science startups.

GPT-5.5 Takes the Lead in Utilization Efficiency, DeepSeek V4 Pro Wins the Title of Best Cost-Performance! Real-World Cybersecurity Attack and Defense Report on Large Models Released

The reasoning capabilities of large language models in the field of cybersecurity are facing a serious test. Security researcher Kasra Rahjerdi conducted simulated hacker attack tests on mainstream large models by building an APK with core vulnerabilities in book review data, revealing their true level of security reasoning and vulnerability exploitation. The test lasted 2 hours with a single budget of $10, intuitively demonstrating the performance of each model in complex logical challenges.

Anthropic Updates Recruitment Guidelines: Prohibits Use of AI Tools in On-Site Interviews, with Annual Salaries Up to $850,000

Anthropic updated its recruitment guidelines, prohibiting candidates from using AI tools during on-site interviews to assess genuine thinking and real-time problem-solving abilities. This move has drawn industry attention, marking a shift in talent screening by AI companies toward deeper and more pure evaluation. The interview process consists of five rounds, with the "culture interview" being the most challenging, requiring candidates to deeply respond to values, worldviews, and ethical issues.