Pulse

Anthropic co-founder Jack Clark predicts a 60% chance that AI systems will be capable of autonomously developing their successors by the end of 2028. This projection is based on rapid advancements in AI's ability to handle tasks like coding, model tuning, and routine experiments. The development suggests AI could soon automate much of its own R&D, though strategic direction and taste remain human-driven. Separately, a demo showcased the creation of a planetary exploration app using GPT Images 2 and Gemini 3.1 Pro, highlighting the creative potential of generative models in application development. AI

Google has reportedly integrated a 4GB Gemini Nano LLM into the Chrome browser without explicit user consent or disclosure, leading to performance issues and increased disk usage for billions of users. This move mirrors concerns raised about Apple's Siri incorporating local LLMs and Microsoft's Copilot integration, sparking debate about user control and the environmental impact of widespread, unprompted AI deployment. Critics argue that such installations should be opt-in, allowing users to manage AI components like any other application, rather than being silently embedded and potentially consuming significant energy resources. AI

AI infrastructure operator IREN has acquired Mirantis for $625 million in an all-stock deal to enhance its AI cloud platform. The acquisition aims to address the challenge of converting deployed GPUs into usable, revenue-generating services by adding Mirantis's Kubernetes management and enterprise support capabilities. This move allows IREN to move beyond providing raw compute power to offering a more integrated software and services layer for AI workloads. AI

OpenAI has developed a new networking protocol called MRC (Multipath Reliable Connection) to address performance bottlenecks in large-scale AI training. This protocol, released through the Open Compute Project, spreads data transfers across hundreds of paths simultaneously to prevent congestion and recovers from failures within microseconds. MRC aims to significantly reduce GPU idle time and improve the efficiency of training massive AI models, potentially increasing training speeds by up to 40%. AI

The increasing demand for AI data centers is causing a global shortage of memory chips, a phenomenon dubbed "RAMageddon." This shortage is driving up the cost of essential components, leading manufacturers like Apple, Microsoft, and Sony to increase prices on products ranging from laptops and smartphones to gaming consoles and VR headsets. Analysts predict that the era of affordable entry-level electronics may end by 2028, as the production capacity for these chips is already committed to AI infrastructure for several years. AI

Mistral AI has released Medium 3.5, a new open-weight model featuring 128 billion parameters and a 256,000 token context window. This model supports multimodal input and adjustable reasoning capabilities. The weights for Medium 3.5 are available on HuggingFace under a modified MIT license, superseding previous models like Mistral Medium 3.1. AI

The Model Context Protocol (MCP) is emerging as a crucial standard for connecting AI agents to external tools and data sources, aiming to simplify integration and reduce development time. Initially an internal experiment at Anthropic, MCP is designed to act as a universal adapter, akin to USB-C for AI, allowing agents to discover and execute tools without custom code for each integration. By providing a standardized way for AI to access real-world data and functionalities, MCP is projected to significantly accelerate agent development and enable more complex, reliable business applications. AI

Google announced Gemini 3.5 Pro at Google I/O, a significant advancement in AI capabilities. Separately, a breakthrough in AI intelligence called SubQ was introduced. AI

Apple's iPhone 17 series dominated the global smartphone market in the first quarter of 2026, capturing the top three sales positions and contributing significantly to Apple's record revenue. The report from Counterpoint Research indicates a high concentration of sales within the top ten models, which accounted for 25% of all shipments. Concurrently, there's a surge in demand for Chinese MCU chips from overseas AI power and optical communication companies to support expanding AI capabilities. AI

Animoca Brands has launched a $10 million fund focused on agentic AI technologies, aiming to foster innovation in decentralized AI. Meanwhile, AMD CEO Lisa Su anticipates a significant surge in the AI-driven server CPU market, projecting a revenue opportunity exceeding $120 billion. AI

DeepSeek V4 was released on April 24, 2026, with its architecture featuring five key tricks that contribute to its cost-effectiveness. The model achieved a Codeforces rating of 3,206, marking it as the highest rating ever recorded. AI

SAP announced its intention to acquire the German AI startup Prior Labs for an undisclosed amount, with plans to invest approximately $1.16 billion over the next four years to develop it into an AI lab focused on structured data. Prior Labs, founded 18 months ago, specializes in tabular foundation models, which are seen as a potentially better fit for enterprises than language models. SAP is also implementing a policy to restrict unauthorized AI agents from accessing its products, while endorsing specific architectures like Nvidia's NemoClaw through its own Joule Agents beta program. AI

Altara, a startup founded by former Harvard computer science students Eva Tuecke and Catherine Yeo, has raised $7 million in seed funding to develop an AI layer for the physical sciences. The platform aims to consolidate scattered technical data from sectors like battery and semiconductor manufacturing, which is often lost in spreadsheets and legacy systems. By analyzing sensor logs, failure reports, and other data, Altara's AI can reportedly condense weeks of manual data triaging into minutes, accelerating product improvement and failure diagnosis. AI

Altara AI has secured $7 million in funding to address data fragmentation and inefficiencies in physical sciences research. The company's platform aims to reduce errors caused by reliance on spreadsheets and legacy systems, which currently hinder scientific data integrity and R&D acceleration. This funding will support Altara's efforts to unify disparate scientific data and improve the accuracy of research outcomes. AI

AMD anticipates a significant downturn in its gaming revenue for the latter half of the year, projecting a decline of over 20%. This forecast is attributed to escalating memory and component costs impacting consumer demand. Despite this, the company expects substantial year-over-year growth in its data center segment, driven by increased AI demand. AI

Theoretical physicist Alex Lupsasca has found that recent AI models, specifically GPT-5.5, are significantly advancing the frontier of scientific research, particularly in physics. While initial public reception to GPT-5.5 was lukewarm due to its perceived lack of improvement in everyday tasks like email writing, Lupsasca observed that its capabilities in complex problem-solving have dramatically expanded. He demonstrated this by having GPT-5.5 reproduce a year's worth of his research in mere minutes, a feat that highlights AI's potential to accelerate scientific discovery. AI

OpenAI plans to invest approximately $50 billion in computing infrastructure for 2025, aiming to fuel the development of advanced AI models like GPT-5.2 and potentially achieve Artificial General Intelligence (AGI). This significant expenditure highlights the escalating competition within the AI industry and the immense computational resources required for cutting-edge AI development. AI

Indium Phosphide (InP) is a critical semiconductor material used in datacenter laser chips and optical transceivers that connect GPUs in AI clusters. Its unique crystal lattice allows for the growth of alloys that emit light at specific wavelengths essential for minimizing signal loss and dispersion in optical fibers. The production of InP is challenging due to slow crystal growth, low yields, and significant capital investment, with new capacity taking years to establish. AI

Anthropic has released a new AI model named "Everyday." The details surrounding this release, including its specific capabilities and intended applications, are not yet widely known. This announcement marks another step in Anthropic's ongoing development of advanced AI systems. AI

Developers are reporting a significant decline in performance with Anthropic's Claude Opus 4.7, leading many to revert to the previous version, Opus 4.6. Users cite issues such as the model arguing with instructions, getting stuck in repetitive reasoning loops, and a drastic drop in performance on long-context retrieval tasks. These regressions appear despite Anthropic's published benchmarks showing improvements, suggesting a potential trade-off in model efficiency or precision for broader multilingual capabilities or to meet demand. AI

FIS and Anthropic have partnered to develop an AI agent aimed at revolutionizing financial crime detection. This collaboration moves beyond typical chatbot applications, focusing instead on autonomous agents capable of executing complex financial operations with minimal human oversight. The initial application, the Financial Crimes AI Agent, can autonomously detect, analyze, and flag suspicious transactions, significantly reducing investigation times and operational costs. The partnership emphasizes deep integration, with Anthropic engineers working directly within FIS's systems to ensure the agents can handle real-world banking complexities and regulatory requirements. AI

Anthropic's new AI model, Mythos, has generated significant concern due to its advanced capabilities in discovering software vulnerabilities. While Anthropic suggests Mythos could revolutionize cybersecurity by enabling comprehensive defense, its potential for misuse has led to widespread apprehension. CEOs Jamie Dimon and Dario Amodei acknowledged the real and escalating risks, with Amodei noting a dramatic increase in identified vulnerabilities compared to previous models, emphasizing the urgent need for rapid patching. AI

China is intensifying efforts to localize its semiconductor supply chain by aiming for 70% domestic silicon wafer usage among its chipmakers this year. This initiative is driven by the increasing importance of chips for AI development and ongoing export restrictions from other nations. While China is largely self-sufficient in older 8-inch wafers, the push is now focused on advanced 12-inch wafers, with companies like Xi'an Eswin Material Technology significantly ramping up production. AI

Nvidia CEO Jensen Huang stated that China should not have access to the company's latest AI GPU technologies, including the Blackwell and Rubin lines. He believes the United States should maintain a lead in AI hardware, advocating for the U.S. to possess "the first, the most, and the best" in this domain. Huang's comments, made at the Milken Institute Global Conference, suggest that restricting China's access to advanced AI accelerators is crucial for maintaining America's technological edge. Despite previous U.S. government approvals for certain Nvidia chips to be sold to China, the company has recently indicated no shipments of its H200 processors to the region. AI

Apple is reportedly in preliminary discussions with Intel and Samsung to diversify its chip manufacturing, seeking US-based production facilities. This move comes as Apple faces constraints from its primary supplier, TSMC, and aims to mitigate risks associated with relying on a single source. The potential partnerships could bolster Intel's foundry business and Samsung's manufacturing capabilities, while reducing Apple's supply chain vulnerabilities. AI

JSR, a major photoresist manufacturer, is constructing its first production facility in Taiwan, with an anticipated operational start by 2028. This move aims to enhance collaboration and co-development of advanced photoresists with TSMC, the world's largest contract chip manufacturer. The new plant is intended to close a competitive gap with rivals Tokyo Ohka Kogyo and Shin-Etsu Chemical, who already have similar facilities in Taiwan. AI

Panthalassa, a startup focused on renewable energy and ocean technology, has secured $140 million in Series B funding. The investment, led by Palantir co-founder Peter Thiel, will support the development of offshore data centers powered by wave energy. These autonomous nodes aim to address the growing electricity demands of AI by providing a sustainable power source and utilizing ocean cooling. AI

Astera Labs has introduced the Scorpio X, a new AI fabric switch designed as an alternative to Nvidia's NVSwitch. This ASIC offers 320 lanes of PCIe 6.0 connectivity and 5.12 TB/s of bidirectional bandwidth, aiming to provide high-speed interconnects for AI systems without requiring proprietary NVLink compatibility. The Scorpio X is intended to facilitate the use of diverse accelerators and enable disaggregated inference architectures by offering a vendor-agnostic solution. AI

Anthropic's Claude 4.5 Opus is reportedly on par with the openly available Qwen3.6-27B model, surpassing its own previous generation. Separately, Google has made its "Ask Gemini in Drive" feature generally available, now supporting Japanese language input. AI

Kevin O'Leary's proposed 40,000-acre AI and cloud computing campus in Utah has received approval from the Box Elder County Commission. The massive project, dubbed 'Wonder Valley,' is expected to consume up to 9 gigawatts of power, potentially doubling the state's current electricity usage. Despite significant local opposition and concerns about environmental impact, O'Leary has dismissed the protests as orchestrated and potentially AI-generated. AI

Anthropic is launching a $1.5 billion venture with Wall Street firms to integrate its Claude AI model across financial services. Separately, the NSA is reportedly testing Anthropic's Mythos AI model for identifying security vulnerabilities in Microsoft products. This testing has prompted Google, xAI, and Microsoft to voluntarily submit their own AI models for national security review. AI

Five major book publishers and author Scott Turow have filed a class-action lawsuit against Meta Platforms and CEO Mark Zuckerberg, alleging the illegal use of millions of copyrighted works to train Meta's Llama AI models. The lawsuit claims Zuckerberg personally authorized and encouraged this infringement, which plaintiffs describe as one of the largest copyright violations in history. Meta has stated it will aggressively fight the lawsuit, asserting that training AI on copyrighted material can be considered fair use. AI

GUC and Wiwynn are collaborating to develop AI infrastructure solutions tailored for hyperscalers, focusing on integrated SoC design, optical I/O, and liquid cooling. Separately, the Michigan Utility Board has imposed a 12-month moratorium on a proposed $1.25 billion data center project by the University of Michigan and Los Alamos National Laboratory, citing concerns over its significant water and power consumption. AI

Google is reportedly preparing to release an updated version of its Gemini Flash model, potentially named Gemini 3.x Flash. This upcoming release is suggested by developers observing new signals and the phased deprecation of the previous Gemini 2 Flash model. The launch is anticipated to occur around the time of Google's I/O 2026 conference. AI

JPMorgan Chase CEO Jamie Dimon publicly endorsed the significant AI capital expenditure, projecting it to exceed $1 trillion next year and stating it's worth the investment. Anthropic CEO Dario Amodei discussed their new model, Mythos, which has raised concerns about cybersecurity vulnerabilities. Amodei clarified that Mythos's limited release was not due to compute constraints, and both leaders expressed optimism that this critical juncture could lead to positive global outcomes. AI

Netomi, a company focused on AI for customer service, has secured $110 million in funding. The investment round saw participation from major players like Accenture and Adobe, signaling their confidence in Netomi's technology. This funding is expected to accelerate the development and deployment of AI solutions aimed at enhancing customer interactions. AI

Voice AI company ElevenLabs has announced new investors as part of its $500 million Series D funding round, including major institutions like BlackRock and enterprises such as NVIDIA and Salesforce. The company also revealed it has surpassed $500 million in annual recurring revenue and its valuation has grown to $11 billion. ElevenLabs aims to develop human-level AI voice models and has recently acquired a Polish voice AI startup to enhance its research team. AI

CopilotKit has secured $27 million in Series A funding to enhance the development of AI agents within applications. The company's AG-UI protocol allows these agents to interact with user interfaces and perform actions, moving beyond simple text-based chatbots. This funding round was led by Glilot Capital, NFX, and SignalFire, and the AG-UI protocol has gained support from major cloud providers like Google, Microsoft, Amazon, and Oracle. AI

India's first generative AI unicorn, Krutrim, is shifting its focus from developing sovereign AI models to offering cloud services by 2026. This pivot is driven by the economic realities and significant GPU shortages impacting the scaling of homegrown AI infrastructure. The move reflects broader challenges faced by nations aiming for technological independence in the AI domain. AI

Google is reportedly planning a new AI model called 'Google AI Ultra Lite,' also known as 'Neon,' and is integrating Gemini into its Chrome browser. Separately, MUFG and Google are forming a strategic partnership to achieve autonomous finance using AI agents. Additionally, Google's Fitbit is launching a new device, 'Fitbit Air,' which will leverage Gemini for health management. AI

Anthropic has released Claude Opus 4.7, a new model described as obedient, discerning, and creative in its responses. The release highlights advancements in the model's ability to follow instructions and exhibit creative capabilities. This update positions Claude Opus 4.7 as a significant step forward in AI development. AI

The burgeoning demand for storage drives by AI data centers is causing significant price increases and shortages, impacting organizations like the Internet Archive and Wikimedia Foundation. These entities are finding it more expensive and difficult to acquire the necessary hard drives and SSDs to continue their work of archiving the internet. This situation is compounded by stricter anti-scraping measures on websites, which inadvertently hinder archival efforts. AI

Anthropic has partnered with financial giants Blackstone and Goldman Sachs to form a $1.5 billion joint venture focused on enterprise AI consulting, aiming to integrate Claude AI into business operations. Meanwhile, OpenAI is facing a $30 billion lawsuit from Elon Musk, who alleges the company abandoned its nonprofit mission for profit, with CEO Sam Altman defending the restructuring. Separately, a trend known as 'job hugging' is emerging, where workers remain in their current roles due to economic uncertainty, despite underlying dissatisfaction. AI

Samsung Electronics has surpassed a $1 trillion market valuation, becoming the second East Asian company after TSMC to reach this milestone. This surge is attributed to booming demand for AI-related stocks and potential collaborations with Apple. The company is reportedly in talks with Apple to manufacture main processors as an alternative to TSMC, driven by supply chain diversification needs and the high demand for HBM chips used in AI data centers. AI

The increasing energy demands of AI models are straining global power grids, leading to concerns about sustainability and capacity. A new study highlights that Europe's grid infrastructure is already facing challenges, with long queues for new facilities and existing ones operating at reduced capacity. Specialized control systems are being developed to manage the fluctuations caused by AI data centers and prevent disruptions to the power supply. AI

Taiwan is establishing a "Quantum National Team" to advance its quantum computing capabilities, aiming to become a leader in the field similar to its dominance in semiconductor manufacturing with TSMC. This initiative is spearheaded by President Tsai Ing-wen and Vice President Lai Ching-te, with support from various research institutions and government bodies. The team's focus includes developing quantum chips, communication technologies, and fostering a domestic quantum industry to enhance national security and economic competitiveness. AI

Gensyn, an AI infrastructure company, has secured $78 million in funding led by Andreessen Horowitz (a16z). The company focuses on the challenge of verifying decentralized AI computations without requiring them to be re-executed. Gensyn's model involves burning 70% of its revenue through its $AI token and is developing prediction markets managed by AI agents. AI

Norway has joined the 'Pax Silica' initiative, a coalition aimed at countering China's growing technological influence. This move signifies a broader international effort to secure critical mineral supply chains and promote technological sovereignty. The initiative seeks to build alliances that can safeguard against economic coercion and ensure stable access to essential resources for advanced technologies. AI

Venture capital firm Y Combinator's stake in OpenAI is now valued at over $5 billion, representing approximately 0.6% of the AI company. This valuation is based on OpenAI's reported $852 billion market capitalization. The significant financial return highlights the transformative impact of AI investments on the venture capital landscape. AI

A new company founded by former Meta AI scientists has secured $1.03 billion in funding. The startup is reportedly exploring alternative methods to Large Language Models (LLMs). This significant investment suggests a strong belief in the company's innovative approach to AI. AI