Discover
Language
Account
Anthropic has unveiled new features for its Claude AI assistant, specifically designed for financial analysts. The update includes an Excel plugin, real-time data connectivity, and specialized AI agent skills, positioning Claude as a competitor to tools like Microsoft Copilot and ChatGPT.
MiniMax has launched its M2 inference model, boasting 230 billion parameters and speeds of 100 tokens per second. Designed specifically for smart agents, the model emphasizes efficiency and real-time interaction capabilities while reducing context window size for practical deployment.
Zoom CEO Eric Yuan forecasts AI will revolutionize workplaces, reducing the traditional five-day workweek to three or four days within five years. He envisions AI assistants handling routine tasks, freeing employees for strategic work.
AI startup Mercor has raised $350 million in a funding round led by Felicis Ventures, reaching a $10 billion valuation. The funds will expand talent networks, upgrade matching systems, and improve service delivery speed.
Fitbit launches its Gemini-powered AI Health Coach, offering personalized fitness and sleep guidance for Premium subscribers. Android users in the U.S. gain early access tomorrow, with iOS support coming later this year.
China's DeepSeek model achieved a 10.61% annualized return in an autonomous AI trading competition hosted by HKU, surpassing major global models and the Nasdaq 100 benchmark. The experiment demonstrated AI's potential in financial decision-making without human intervention.
OpenAI has called for increased US power infrastructure investment to maintain its AI leadership over China. CEO Sam Altman emphasized electricity as a strategic asset, citing China's significant lead in power capacity expansion.
Anthropic has unveiled a finance-specific version of Claude AI, featuring direct Excel integration, real-time market data access, and advanced financial analysis tools. The upgrade reportedly reduces analysts' workloads by 80%, transforming how financial professionals interact with data and models.
The U.S. Department of Energy and AMD have announced a $1 billion partnership to develop two advanced supercomputers aimed at accelerating breakthroughs in nuclear energy, healthcare, and national security. The systems, named 'Lux' and 'Discovery,' will leverage AMD's AI chips to tackle complex scientific challenges.
European AI startup Mistral has launched Mistral AI Studio, a production platform enabling enterprises to build, monitor, and scale AI applications. The platform focuses on governance, observability, and agent runtime while offering EU-based infrastructure and multimodal capabilities.
Tahoe Bio has launched Tahoe-x1 (Tx1), a 3B-parameter AI model designed to decode gene-cell-drug interactions in cancer research. The model outperforms existing tools in identifying cancer driver genes, reconstructing carcinogenic pathways, and predicting drug efficacy, potentially accelerating personalized therapies.
Qualcomm has launched its next-generation AI chips, AI200 and AI250, targeting NVIDIA's dominance in the market. The announcement spurred a 20% surge in Qualcomm's stock price. The chips feature advanced technologies like liquid cooling and near-memory computing, aiming to reduce costs and boost performance for AI inference tasks.
Pinterest has introduced AI-powered upgrades to its Boards feature, including personalized outfit collages and curated shopping recommendations. These updates aim to transform Pinterest into an AI-driven shopping assistant while maintaining transparency around AI-generated content.
OpenAI's GPT-5 demonstrates significant improvements in handling mental health queries, reducing unsafe responses by 65%. Collaborating with 300 global experts, the model now better addresses suicide risks, psychosis, and emotional dependency. Compliance in suicide-related tests rose to 91%, marking a leap in AI safety.
German tax consultancy Steuerrecht.com leverages ChatGPT Business to streamline operations, enhance client services, and maintain GDPR compliance. The AI solution accelerates contract processing, improves legal research, and powers multilingual communication while safeguarding sensitive data.
Biotech startup Nephrogen, founded by Demetri Maxim, combines AI and CRISPR technology to develop a revolutionary kidney disease treatment. Their drug delivery system claims 100x efficiency over FDA-approved methods, with clinical trials planned for 2027.
AI startup Thinking Machine has unveiled On-Policy Distillation, a novel training method that enhances small model efficiency by 50-100 times. The technique combines reinforcement and supervised learning, drawing praise from OpenAI's former CTO Mira Murati. Early tests show dramatic improvements in mathematical reasoning tasks while solving 'catastrophic forgetting' challenges.
AI startup Mercor raised $350 million in a funding round led by Felicis, accelerating its pivot to AI data annotation services following Scale AI's reduced market presence. The company now manages 30,000 contract workers and faces growing competition from Surge AI and Turing AI.
Feimao's new AI WiFi simultaneous interpretation device addresses two major challenges in international tech collaborations: unstable networks and language barriers. Combining reliable connectivity with offline translation capabilities supporting 11 languages, the system enhances cross-border meeting efficiency with real-time translations and automated multilingual minutes.
NVIDIA has introduced OmniVinci, a cutting-edge multimodal understanding model that outperforms existing benchmarks by 19.05 points while using just one-sixth of the training data. The model integrates vision, audio, and text processing into a unified system, showcasing exceptional efficiency and performance.
