Discover the AI Tools

Anthropic, the AI company behind Claude, is gearing up for an IPO with staggering numbers. Its potential market could hit $30 trillion, dwarfing SpaceX's record. With annualized revenue already at $65 billion and projections of $200 billion by 2028, the company is positioning itself as a tech giant. But can it live up to the hype? Let's dive into the details.

OpenAI has developed its own AI chip, Jalapeno, which reportedly surpasses NVIDIA's GB300 in performance tests. This move signals a push for hardware independence and could reshape the AI infrastructure landscape. The chip shows superior efficiency and speed, promising lower costs and better performance for AI tasks.

Stability AI, the company behind Stable Diffusion, has raised $76 million in Series B funding, with investments from Universal Music, Sony Music, Warner Music, and EA. This strategic move signals a deeper collaboration between AI and the music industry. The funds will boost their creative tools and professional services, while a legal win against Getty Images in the UK eases some pressure.

Google has unveiled Gemini Enterprise for Legal, a suite of AI tools tailored for law firms, intensifying competition in the legal tech space. The platform integrates with existing legal software and offers AI agents to handle routine tasks, letting lawyers focus on higher-value work. This move follows similar efforts by rivals like Anthropic and Thomson Reuters, signaling a shift from model competition to workflow integration.

阿里巴巴创始人马云近期通过新股配售增持公司港股超6亿港元，加上蔡崇信、吴泳铭的连续增持，管理层累计增持超8亿港元。同时，阿里宣布80亿港元配股计划，资金将全部投入AI建设，获近三倍超额认购，显示市场对阿里AI转型的信心。

Anthropic has upgraded Claude's memory system, allowing the AI assistant to remember details across its chat and Cowork features. This means no more repeating yourself when moving from brainstorming to execution. The update, rolling out to all users, also includes transparent controls for managing what Claude remembers, addressing privacy concerns head-on.

Anthropic has rolled out a major upgrade to Claude's memory system, breaking down the wall between Cowork and regular chats. Now, what Claude remembers from your daily conversations can be used in Cowork tasks, and vice versa. This means less repeating yourself and smoother workflows. The update also brings thoughtful privacy defaults, keeping sensitive info out of memory unless you opt in. Available across all plans and platforms, this change makes Claude feel more like a true assistant that actually remembers.

Tencent has hit back at reports claiming its gaming division is overhauling its AI operations. The company says the story is riddled with inaccuracies, especially the part about creating a new department. Instead, it clarifies that the Game AI Engine Department has always been part of its public tech line, focused on boosting efficiency and innovation. This isn't about expansion—it's about doubling down on practical results.

SpaceX's $6 billion acquisition of Cursor marks a pivotal moment in AI. Musk's candid admission of lagging behind rivals like Anthropic, coupled with a push for aggressive integration, signals a high-stakes race for AI dominance. This deal isn't just about money—it's about securing the future of AI development.

Anthropic has unified memory across Claude's Cowork and chat features, so you no longer need to repeat yourself. The AI remembers your context from one place and applies it to the other, making workflows smoother. Sensitive info stays off by default, but you can switch it on anytime. Here's what this means for your daily AI use.

Apple's latest Mac lineup, featuring the M6 chip and the monstrous M5 Ultra, is set to redefine desktop AI. With the ability to run trillion-parameter models locally and a 4.8x speed boost over the M4, these machines are a game-changer. But the power comes at a price—literally. Starting at $899 for the Mac mini, the new lineup is pricier, yet for AI enthusiasts, it might just be worth every penny.

NVIDIA is making a bold move in the AI race, investing $6 billion to build one of the world's most powerful open-source models. Teaming up with startup Poolside, the chip giant aims to challenge the likes of DeepSeek and OpenAI. With a $1 billion stake in Poolside and over 100 engineers assigned to the Nemotron project, NVIDIA is signaling that open source is the future—and it's ready to lead the charge.

China's aerospace sector just took a giant leap. The 'Huashan' AI model, developed by Zhongke Tianta, has successfully completed in-orbit verification aboard the 'Chen Guang One' satellite. This marks a shift from ground-controlled satellites to ones that can think and act on their own, potentially revolutionizing how we manage massive satellite constellations.

A leaked report suggests OpenAI has finished pre-training a massive new model called Bel, with over 10 trillion parameters. This could be the foundation for GPT-6 and a stepping stone toward AGI. But what does this mean for the AI race? Let's dive in.

Generalist, a robotics startup founded by ex-DeepMind and Boston Dynamics engineers, has reached a $3 billion valuation after a $200 million funding round led by 8VC. This extension brings its Series B total to $600 million. The company's AI models let robots learn tasks from short video clips, but the industry still faces hurdles in commercialization.

Google has officially entered the legal AI space with Gemini Enterprise for Legal, a tool designed for law firms. This move intensifies competition with other tech giants and AI startups. The platform promises to handle routine and complex legal tasks while ensuring data security. Google collaborated with major law firms to develop the product, which also integrates with specialized legal AI tools like Harvey. As the legal industry increasingly adopts AI, this launch marks a significant step in transforming legal work.

Stability AI has secured $76 million in Series B funding, with investors including game giant EA, three major record labels, and AMD's venture arm. This unusual alliance signals a shift from conflict to collaboration between AI companies and content rights holders. The funds will drive creative product development and applied research.

xAI is rolling out a wave of upgrades for its Grok Bot, turning it into a more versatile tool for both personal and professional use. New features include shareable templates, multi-account switching, Chrome profile import, and a clever exit routing system that bypasses IP blocks. These updates signal a push toward building a robust ecosystem for AI agents, making them more practical for everyday tasks and complex workflows.

Tencent's executive Tang Daosheng addressed criticism about the company's AI pace in an internal post, calling competition a marathon where endurance matters more than speed. He acknowledged internal anxiety over rivals' progress and computing power constraints but emphasized long-term investment and scenario advantages. This echoes CEO Ma Huateng's earlier 'leaking boat' metaphor, signaling Tencent's strategy of patient, steady advancement in AI.

Apple has unveiled the M6 and M5 Ultra chips, marking a significant leap in desktop performance and on-device AI. The M6, built on a 2-nanometer process, delivers up to 13.5 times faster AI prompt processing than the M1. Meanwhile, the M5 Ultra introduces a four-die architecture with massive GPU and memory bandwidth, pushing the boundaries of local AI computing. This move signals a shift toward edge AI, reducing reliance on cloud computing.