Written by: Techub News Compilation

On the eve of the annual Google I/O Developer Conference, Google unexpectedly released a series of significant AI products and strategic collaborations, among which the most notable is its first laptop designed specifically for AI, the "Google Book." This represents not only a hardware innovation but also Google's comprehensive evolution from an "operating system" to an "intelligent system." Meanwhile, Google's moves in AI infrastructure and cutting-edge applications are equally rapid, including a partnership with SpaceX to build a space AI data center and its biotechnology subsidiary, Isomorphic Labs, securing $2.1 billion in funding. These initiatives collectively outline Google's full-stack advantage as it competes in the next generation of AI.

AI Laptop: Google Book and the Smart System Revolution

The "Google Book" laptop released by Google is designed from the ground up for AI. Its core philosophy is to create a "smart system" rather than a traditional operating system. This means AI capabilities, particularly its flagship model Gemini, will be deeply integrated into every interaction layer of the device, becoming the core of the system.

A signature feature of this device is the "Magic Pointer." Users simply hover the cursor over a date in an email, and Gemini will automatically schedule a meeting; hovering over a picture of a living room, Gemini can synthesize a rendering of a new sofa. Users can even ask it to plan a family gathering, and it will automatically create a real-time dashboard containing flight, hotel information, and a countdown. These features are natively integrated into this new device.

From a hardware perspective, the Google Book is a natural evolution of the Chromebook concept. In 2011, Google launched the Chromebook, reshaping the way laptops are used with a browser-centric model. Now, the Google Book represents a new wave of transformation: AI is "consuming" the browser, becoming the new core interaction layer. The device itself has an elegant design, resembling a hybrid of the MacBook Air and MacBook Pro, priced competitively between $200 to $500.

More importantly, the Google Book deeply integrates with the Android phone ecosystem, providing users with a cohesive experience similar to Apple's ecosystem, but delivering AI capabilities in software that Apple promised but failed to deliver. For iOS users, it may seem like an interesting experimental device, but its demonstrated AI-native experience signals the direction of personal computing devices' development.

Ecological Expansion: Eating into Apple's Market and Forming AI Alliances

The release by Google is not just about hardware; it is a comprehensive ecological offensive. They introduced "Gemini Intelligence," a model system for AI that operates across all of Google's applications, tools, and products (such as Gmail, Maps, and G Suite). Google's vertical integration advantage is highlighted here: they own the model layer (Gemini), the compute layer (GPU), and achieve unparalleled distribution capability through a massive product matrix.

Google is actively simplifying the data migration process for users moving from the Apple ecosystem to the Google ecosystem. This reflects Google’s open ethos akin to that of Android. Analysis suggests that as Apple seems to falter in AI progress, Google is seizing the opportunity to eat into its market share. Although Apple signed a multi-billion dollar licensing agreement for the Gemini model with Google, it has not shown signs of building a foundational model itself. Google, seizing the moment, released new products a week before the major presentation, actively competing for market share.

Additionally, an "AI alliance" composed of companies such as SpaceX AI, Anthropic, Tesla, Google, and Cursor is forming. This is a mutually beneficial symbiotic relationship: Google gains an inexpensive pathway to space and unlimited solar energy; Anthropic receives 300 megawatts of inference computing power from SpaceX's "Colossus One" data center; SpaceX generates about $5 to $10 billion in revenue through its dealings with Anthropic and Cursor; Cursor gains access to flagship coding model computing power that they could not afford otherwise. Notably, OpenAI currently appears to be excluded from this alliance.

Space Ambition: Google and SpaceX's AI Data Center Cooperation

This week, another major announcement is Google's partnership with SpaceX in the realm of AI data centers. SpaceX appears to be collaborating with several AI giants to deploy data centers in space. Following its agreement with Anthropic last week, Google joined in this week. SpaceX will utilize its payload capacity to send Google's TPUs (Tensor Processing Units) into space.

This is not a fleeting novelty. Google CEO Sundar Pichai announced approximately six months ago that the company was developing radiation-resistant TPUs for space use. They needed a way to send these devices into orbit, and SpaceX provides the most economical "space highway." Google is already a shareholder in SpaceX (holding 6.1% of the shares), and both have a shared interest base. Google also has the existing space machine learning project "Project Suncatcher" and collaborates with other rocket launch companies and satellite design companies like Planet Labs.

As SpaceX prepares for its IPO, one of its clear goals is to become the infrastructure provider in this new space race. The collaboration between Google and SpaceX marks the formal extension of competition in AI computing infrastructure into the space domain.

Cutting-Edge Biotechnology: Isomorphic Labs and AI-Driven Drug Discovery

Google's subsidiary focused on AI drug discovery, Isomorphic Labs, announced this week that it has secured $2.1 billion in funding, led by Thrive Capital. The company is spearheaded by Demis Hassabis, CEO of Google DeepMind, and is considered the "DeepMind of the biotech field."

Its breakthrough achievements stem from early work on the "protein folding" problem. Proteins are key to regulating human functions, and understanding their structure is critical for treating specific diseases. The AI models developed by Demis Hassabis's team (such as AlphaFold and AlphaGo) can predict the three-dimensional folding structures of proteins and design molecular drugs that can precisely match specific protein "lockholes" like a "key."

The core technology of Isomorphic Labs is a model known as "ISO DDE" (Isomorphic AI Drug Design Engine), which can identify a vast number of new molecules. This technology is freely used by about 300,000 frontier researchers worldwide and has discovered a series of new molecules that may be used to treat major diseases such as Alzheimer's and cancer. This substantial funding will be used to drive these discoveries into human trial stages, with hopes of producing the first practical therapies in the coming years.

This technology not only concerns disease interception; in the long run, it may open the door to manipulating human functions, achieving "sci-fi" abilities such as seeing infrared light with the naked eye. Demis Hassabis's long-standing dedication to this field makes him a key figure in driving this revolutionary product.

Hardware Marvels and China's Robotics Progress

In addition to software and infrastructure, there have also been notable advancements in the hardware field. A giant mecha robot available for "driving" has emerged in China, priced at approximately $50,000. This robot weighs over 500 kilograms and even demonstrated its ability to knock down brick walls during a demonstration. Although its practical use seems ambiguous ("just because it can"), it showcases possible future forms of devices and reflects China's traditional advantages in robotic hardware manufacturing and scaling.

Meanwhile, Thinking Machines Labs, founded by former OpenAI CTO Mira Murati, released its new model this week after nearly two years of silence. This model is not a traditional LLM (Large Language Model) but rather an innovative "interactive model." Its uniqueness lies in its "single-modal" design, capable of simultaneously processing audio, video, and text inputs to achieve real-time, bidirectional interactions similar to human communication. It can listen to user interruptions while speaking and respond instantly, overcoming the "one-way conversation" limitations present in many current AI interactions.

However, this model has approximately 12 billion parameters, relatively small compared to some leading models today (rumored to reach 1.5 trillion parameters). This means its level of intelligence may be limited. Hours later, Meta also released a similar AI voice dialogue product. This reflects that, even with renowned founders and massive funding, small AI labs still face immense challenges in catching up with large AI labs (such as OpenAI and Anthropic).

Investment Market Dynamics and AI Company Equity Disputes

As one of the hottest AI labs currently, Anthropic has raised hundreds of billions in funding while also triggering a secondary market investment controversy. Since its equity is not publicly listed, an active "secondary market" has emerged, where investors with certain allocations can resell their shares to ordinary retail investors through special purpose vehicles (SPVs).

This week, a user named Ash Aurora claimed on social media that he facilitated a secondary market transaction for Anthropic through intermediaries, profiting more than his entire net worth accumulated over two decades of career. This sparked widespread attention. Subsequently, Anthropic quickly updated its support page, stating that "any sale or transfer of shares not approved by the board is invalid and will not be recognized by the company."

OpenAI has also issued similar statements. This means many investors who have invested in these SPVs through unofficial channels may not be able to realize their equity when the company goes public in the future, putting their funds at risk. This move has, to some extent, cooled speculation in the secondary market. At the same time, on the blockchain, Anthropic's equity has been tokenized and traded at valuations as high as $15 trillion, far exceeding its actual valuation (rumored to be around $90 billion in the latest funding round). The company’s official clarification helps recalibrate the market.

Global AI Political Economy: U.S. Business Leaders Visit China

This week, the CEO of SpaceX AI and others arrived in Beijing as part of a U.S. delegation. The delegation includes tech leaders such as Elon Musk and Jensen Huang, with goals involving trade rebalancing, energy security (especially to facilitate Iran's opening of the Strait of Hormuz for a peace agreement), the rare earth supply chain, and bilateral discussions on AI risks and security.

Choosing Musk and Huang to accompany them is not coincidental. Musk's Tesla has long been rooted in the Chinese market, while Huang has been striving to sell NVIDIA GPUs in China. He previously stated that understanding China's AI progress is crucial, and through the hardware they use (especially U.S. hardware), one can infer the level of their AI models.

However, current U.S. policies tend to prohibit selling advanced GPUs to China and demand the "reshoring" of manufacturing and GPU production to the U.S. This puts both sides in a deadlock: China needs to purchase GPUs, but the U.S. intends to restrict them. In response, the Chinese government has ordered its major AI labs to use domestic hardware and GPUs to train models. Recent models such as DeepSeek V4 and Kimi K2 are primarily based on training with domestic GPUs like those from Huawei, and they perform quite well, rivaling Claude Opus 4.7 in some aspects, but at a lower cost and faster speed. This visit may aim to ease tensions and open the massive Chinese market for companies like NVIDIA.

As these tech leaders visit, the global competitive and cooperative landscape in the field of AI is undergoing subtle but significant changes.

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。