DD
MM
YYYY

PAGES

DD
MM
YYYY

spot_img

PAGES

Home Blog Page 12

Cursor Launches Mobile App as AI Coding Shifts From Writing Code to Managing Autonomous Agents

0

Artificial intelligence coding startup Cursor is expanding beyond the desktop with the launch of its first mobile application, betting that the future of software development will increasingly revolve around supervising autonomous AI coding agents rather than manually writing code.

The new iOS app, announced on Monday, allows developers to create, monitor and communicate with Cursor’s AI coding agents directly from their smartphones, extending the company’s push into agentic software development just weeks after its blockbuster $60 billion acquisition by SpaceX.

Rather than replacing the desktop experience entirely, the application serves as a companion platform that enables developers to remain connected to coding projects while away from their computers. Users can launch entirely new coding agents from their phones or continue conversations with agents that were originally started on Cursor’s desktop application.

The launch builds on Cursor 2.0, unveiled in October, which marked a significant shift in the company’s strategy from AI-assisted code completion toward autonomous coding agents capable of independently handling increasingly complex software development tasks.

The AI software development industry is witnessing a transformation. Instead of using AI merely to autocomplete lines of code or answer programming questions, developers are increasingly assigning entire programming tasks to intelligent agents that can write, edit, debug and test software autonomously while human programmers focus on reviewing outputs, setting objectives and making architectural decisions.

This transition is changing not only how software is written but also where developers work. Traditionally, professional software engineers relied on powerful desktop computers equipped with multiple monitors to manage large codebases, documentation, and debugging tools simultaneously. Agentic AI is beginning to reduce that dependency because developers no longer need to constantly interact with every line of source code.

Instead, much of their work involves issuing instructions, reviewing progress and refining objectives for AI agents that execute programming tasks independently.

That workflow lends itself naturally to mobile devices. With coding agents running remotely in the cloud, smartphones increasingly function as control interfaces rather than programming workstations, allowing developers to supervise software projects from virtually anywhere.

Cursor is not alone in pursuing this vision. The company joins a growing list of leading AI developers bringing coding capabilities to mobile platforms. Both Anthropic and OpenAI have introduced mobile applications that enable users to interact with their AI coding assistants, reflecting growing demand for continuous access to AI-powered development tools.

Industry observers are describing this as the emergence of “ambient software engineering,” where AI agents work continuously in the background while developers periodically review progress instead of actively writing code throughout the day.

One of the strongest public endorsements of that shift has come from within Anthropic itself. Speaking during a recent presentation, Anthropic’s Claude Code lead Boris Cherny said the arrival of autonomous coding agents had fundamentally changed his own workflow.

“Most of my coding now is on my phone,” Cherny said. “I would have said ‘you’re crazy’ if you told me that six months ago, but yeah, here we are.”

Developer habits are quickly evolving as AI systems become more capable of independently understanding codebases, planning software changes and executing programming tasks with minimal supervision. The mobile launch also demonstrates that Cursor is continuing to execute its product roadmap despite undergoing one of the largest acquisitions in artificial intelligence history.

SpaceX’s $60 billion acquisition of Cursor signaled Elon Musk’s ambition to integrate advanced AI coding capabilities into the company’s broader software engineering ecosystem, including its work on aerospace systems, robotics, and artificial intelligence infrastructure.

Rather than slowing development, Cursor appears to be accelerating product releases. The company’s strategy aligns with a wider trend across the AI industry, which has seen leading firms competing to build not simply coding assistants but fully autonomous software engineering platforms capable of handling larger portions of the development lifecycle.

As AI models improve their ability to reason through complex programming problems, industry analysts expect developers to spend progressively less time typing code and more time defining business objectives, reviewing AI-generated work and orchestrating multiple specialized coding agents.

Cursor’s mobile application is another step toward that future, where software development becomes less about writing every line of code manually and more about directing increasingly capable AI systems that can build, modify, and maintain software on behalf of human engineers.

Baidu Shares Jump on Report Kunlunxin Could Seek Hong Kong IPO at $50 Billion Valuation

0

Shares of Chinese technology giant Baidu climbed more than 7% in Hong Kong on Monday after a report said its artificial intelligence chip subsidiary, Kunlunxin, is preparing for a Hong Kong initial public offering that could value the business at as much as $50 billion.

The rally comes as Chinese AI chipmakers increasingly attract investor interest amid Beijing’s push to reduce dependence on foreign semiconductor technology and challenge U.S. dominance in the rapidly expanding AI hardware market.

According to The Information, citing two people familiar with the matter, prospective investors have been asked to commit to purchasing semiconductor products worth three to seven times the value of their intended investment in Kunlunxin’s planned share sale, an unusual structure designed to demonstrate commercial demand for the company’s chips while strengthening customer relationships.

The report added that Baidu confidentially submitted a listing application for Kunlunxin to the Hong Kong Stock Exchange earlier this year, although the size, valuation, and final structure of the offering had yet to be determined.

Neither Baidu nor Kunlunxin has publicly commented on the report.

Founded in 2011, Kunlunxin was established to develop artificial intelligence processors primarily for Baidu’s own cloud computing, search, and large language model businesses.

Although Baidu continues to hold a controlling stake, the semiconductor company has increasingly operated as an independent business and has expanded sales to external customers over the past two years, positioning itself as one of China’s emerging domestic AI chip suppliers.

The company develops AI accelerators used for machine learning inference, cloud computing, autonomous driving and generative AI workloads, areas that have become strategic priorities for China following tightening U.S. export restrictions on advanced semiconductor technologies.

ByteDance, the owner of TikTok, had expressed interest in Kunlunxin’s processors as Chinese technology companies diversify chip suppliers amid restrictions on access to leading-edge U.S. AI hardware, according to a Reuters report.

The potential IPO comes as China accelerates investment across the AI value chain, spanning semiconductors, computing infrastructure, cloud services and foundation models.

U.S. export controls have significantly limited Chinese companies’ access to the most advanced AI processors produced by Nvidia and other American firms, prompting Beijing to support domestic alternatives capable of powering increasingly sophisticated AI applications.

That effort has benefited companies including Kunlunxin, Huawei’s semiconductor operations, Cambricon, and other Chinese chip developers seeking to build competitive AI computing platforms. Investor enthusiasm reflects expectations that demand for domestic AI chips will continue rising as Chinese cloud providers, internet companies and industrial users expand deployment of large language models and AI-powered applications.

China Narrowing The Technology Gap

While the United States maintains a clear lead in advanced AI semiconductors, analysts believe China is narrowing the gap.

According to Brussels-based economic think tank Bruegel, the United States currently remains ahead in what it described as the artificial intelligence “hardware stack,” encompassing the advanced semiconductors, computing infrastructure and supporting technologies required to train and deploy frontier AI models.

“Despite Chinese progress, the United States remains for now ahead in the race for dominance over the so-called artificial intelligence hardware stack – the resources and equipment, especially semiconductors, needed to run AI models,” Bruegel said in a recent report.

However, the think tank noted that China’s progress has become increasingly evident.

“The signs of Chinese catch-up are real,” Bruegel added, citing China’s growing open-source AI ecosystem, state-backed research pipeline and a sufficiently large domestic market capable of sustaining AI companies while their technologies mature.

China’s vast home market gives domestic AI hardware developers a significant advantage by providing large-scale commercial deployment opportunities even as international expansion remains constrained by geopolitical tensions.

If completed at a valuation approaching $50 billion, Kunlunxin would become one of the world’s most valuable standalone AI semiconductor companies and one of the largest technology listings in Hong Kong in recent years.

The prospective offering also denotes investor focus from consumer-facing AI applications toward the infrastructure underpinning artificial intelligence, including chips, cloud computing, networking equipment and data centers. That trend has driven soaring valuations for AI hardware companies globally as hyperscalers and technology firms continue investing hundreds of billions of dollars to expand AI computing capacity.

Spinning off Kunlunxin is expected to unlock shareholder value while providing the chipmaker with independent access to capital needed to accelerate research, manufacturing partnerships, and commercial expansion.

Tencent Signs $2.9bn Memory Deal With CXMT as China Accelerates AI Chip Self-Reliance Ahead of Landmark IPO

0

Chinese memory chipmaker ChangXin Memory Technologies (CXMT) has secured a long-term supply agreement worth more than 20 billion yuan ($2.94 billion) with internet giant Tencent, in a deal that underscores China’s accelerating push to build a self-sufficient semiconductor ecosystem as demand for artificial intelligence infrastructure continues to surge.

According to people familiar with the matter who spoke to Reuters, the agreement covers the supply of dynamic random-access memory (DRAM) chips for Tencent’s servers over several years. Two sources said the contract runs for as long as three years, while another said it could extend to five years.

The previously undisclosed agreement comes at a pivotal moment for the Hefei-based semiconductor manufacturer as it prepares for one of mainland China’s largest stock market debuts in recent years. In May, the Shanghai Stock Exchange approved CXMT’s application to list on the STAR Market, where the company plans to raise 29.5 billion yuan to finance further expansion.

The size of Tencent’s commitment represents one of the strongest commercial endorsements yet for China’s largest DRAM producer, signaling growing confidence among the country’s biggest technology companies in domestically produced memory chips at a time when geopolitical tensions continue to reshape global semiconductor supply chains.

The agreement is also part of a bigger shift by China’s leading cloud and artificial intelligence companies, which are increasingly seeking to reduce dependence on foreign semiconductor suppliers as U.S. export restrictions continue to limit China’s access to advanced chip technologies.

DRAM chips are indispensable components in modern data centers, powering cloud computing platforms, artificial intelligence training clusters, enterprise databases and high-performance computing systems. Unlike processors that execute calculations, DRAM temporarily stores the massive volumes of data AI models require to operate efficiently. As AI models become larger and more computationally intensive, memory capacity has emerged as one of the industry’s most critical bottlenecks.

The global shortage of advanced memory chips has made long-term supply agreements increasingly valuable.

While details of the contract remain confidential, it was not immediately clear whether the agreement includes CXMT’s high-bandwidth memory (HBM), the premium category of memory used alongside advanced AI processors from companies such as Nvidia and AMD. HBM has become one of the most strategically important components in AI servers because it significantly increases processing speed while reducing power consumption.

Founded in 2016 with strong backing from Chinese state funds, CXMT has become the centerpiece of Beijing’s efforts to establish an indigenous DRAM industry capable of competing with global leaders Samsung Electronics, SK Hynix, and Micron Technology.

The company has long trailed its South Korean and American rivals technologically. However, China’s determination to strengthen semiconductor self-sufficiency has accelerated investment, research, and customer adoption, particularly as export controls imposed by the United States have restricted Chinese companies’ access to advanced foreign semiconductor technologies.

Tencent’s decision to secure multi-year supplies from CXMT illustrates how China’s largest internet companies are increasingly integrating domestic chipmakers into their long-term AI infrastructure strategies.

The company is not alone.

According to additional sources familiar with the discussions, CXMT is negotiating similar long-term collaborations with several other major Chinese technology companies. Its IPO prospectus already identifies Tencent, Alibaba Cloud, ByteDance, Lenovo, and Xiaomi among its major customers.

These partnerships are becoming more important as Chinese hyperscale cloud providers rapidly expand artificial intelligence infrastructure to compete with global leaders such as Amazon Web Services, Microsoft Azure and Google Cloud.

The DRAM Market is Having A Field Day

The timing of the agreement also coincides with one of the strongest memory market upcycles in years.

According to UBS, DRAM contract prices surged approximately 95% quarter-on-quarter during the first quarter of 2026 as AI-driven demand collided with constrained supply. The investment bank expects the memory supercycle to continue until at least the end of 2027.

UBS estimates the global memory market could grow to $786 billion this year before expanding further to approximately $1.2 trillion in 2027, fueled largely by investments in generative AI, cloud computing and high-performance data centers.

That favorable pricing environment has dramatically transformed CXMT’s financial performance. The company reported first-quarter revenue of 50.8 billion yuan, representing an extraordinary 700% increase from a year earlier. It also swung to a net profit of 25 billion yuan after recording a loss of 1.6 billion yuan during the same period last year, highlighting how rising memory prices and surging AI demand have reshaped industry economics.

Large cloud computing companies have increasingly responded to supply shortages by locking in future production through multi-year purchasing agreements.

UBS noted that contracts featuring price bands, advance payments and guaranteed purchase volumes have become common across the industry, with some hyperscalers already committing more than half of their projected memory requirements over three-to-five-year periods.

For CXMT, the Tencent agreement provides both predictable revenue and validation ahead of its public listing. The company is simultaneously embarking on one of the largest capacity expansion programmes in China’s semiconductor industry.

Sources familiar with the plans said CXMT has begun constructing a new DRAM manufacturing facility in Shanghai in addition to its existing packaging plant dedicated to high-bandwidth memory products. The company currently operates three 12-inch wafer fabrication plants, including two in Hefei and one in Beijing, with a combined production capacity of roughly 300,000 wafers per month.

Once the new Shanghai facility and other planned expansions become operational, CXMT expects to double production capacity to approximately 600,000 wafers each month, significantly strengthening China’s domestic memory manufacturing capability.

The expansion reflects Beijing’s broader industrial strategy of building resilient domestic semiconductor supply chains capable of supporting future growth in artificial intelligence, advanced manufacturing and cloud computing while reducing exposure to foreign technology restrictions.

Nevertheless, significant technological challenges remain. According to one of the sources familiar with the company’s operations, CXMT encountered relatively low production yields for its next-generation DDR5 memory products during the first quarter, indicating that the company still trails established competitors in manufacturing efficiency and advanced process technology.

Industry analysts note that while China’s memory industry has made remarkable progress in expanding capacity, closing the technology gap with Samsung Electronics and SK Hynix remains a longer-term challenge requiring sustained investment in research, manufacturing expertise, and advanced process engineering.

Even so, the Tencent agreement signals that commercial acceptance of Chinese-made memory chips is accelerating.

BT, Verizon Form $4 Billion Global Enterprise Joint Venture in Strategic Telecoms Shake-Up

0

BT and Verizon have agreed to combine their international enterprise businesses into a 50:50 joint venture, creating a global communications provider with approximately $4 billion in annual revenue that will target multinational corporations seeking secure cross-border connectivity, cloud networking and managed digital services.

The deal marks one of the most significant restructurings in the international telecom services market in recent years and reflects a broader industry trend toward consolidation as operators struggle with slowing growth, rising infrastructure costs and intensifying competition from cloud providers and specialist networking firms.

Under the agreement announced on Monday, Verizon will make a $625 million equalization payment to BT, while both companies will retain equal ownership and voting rights in the new venture.

The combined business will serve more than 3,000 multinational customers across over 180 countries, bringing together complementary customer bases, global network assets and enterprise service portfolios aimed at large corporations operating across multiple jurisdictions.

Creating A Larger Global Enterprise Player

The joint venture is designed to strengthen both companies’ positions in the highly competitive international enterprise communications market, where customers increasingly demand integrated networking, cybersecurity, cloud connectivity and managed digital infrastructure across global operations.

Verizon Chief Executive Dan Schulman described the partnership as a strategic response to changing customer requirements.

“The venture was the clear answer for international customers who need secure, flexible connectivity that works across borders and cloud environments,” he said.

BT Chief Executive Allison Kirkby said the international enterprise market remains highly fragmented, suggesting the new company could become a platform for broader industry consolidation.

“This is a very fragmented market and this could be the start of further consolidation,” Kirkby told Reuters. “We could possibly look to bring in third parties at some point in the future.”

Her comments indicate the venture could eventually expand through acquisitions or partnerships with additional telecom operators seeking greater scale in international enterprise services.

The transaction represents a major milestone in Kirkby’s broader restructuring of BT since taking over leadership of the 180-year-old British telecommunications group.

Since becoming CEO, Kirkby has focused on simplifying BT’s operations, strengthening its domestic UK business, and reducing exposure to slower-growing international operations.

BT’s international division has weighed on group earnings for several years as enterprise customers shifted spending toward software-defined networking, cloud computing, and integrated digital infrastructure, increasing competition from both traditional telecom rivals and technology companies.

Media reports last month suggested BT had revived discussions with several potential partners for its international business, including AT&T, Orange, and Verizon. The agreement with Verizon, therefore, represents the culmination of a lengthy strategic review aimed at improving the competitiveness of BT’s international operations while allowing management to concentrate more resources on its UK fiber broadband and mobile businesses.

Limited Customer Overlap

Kirkby said the two companies’ customer bases are largely complementary rather than overlapping, increasing the strategic value of the transaction.

“We see this as a unique opportunity to create a scaled player to serve our multinational customers much better,” she said.

She added that there was only limited overlap between existing enterprise clients, allowing the combined business to broaden its addressable market without substantial customer duplication.

The venture is expected to provide multinational companies with broader international network coverage, integrated cybersecurity capabilities, cloud networking services, and managed communications across Europe, North America, Asia-Pacific, and emerging markets.

The companies have appointed Martijn Blanken as chief executive-designate of the joint venture. Blanken, who previously held senior executive roles at Telstra in Australia and Dutch telecommunications operator KPN, will join BT Group on September 1, 2026, before assuming leadership of the combined business as preparations continue for the venture’s launch.

His appointment brings extensive experience in managing large-scale international telecommunications businesses undergoing digital transformation.

Financial Impact

The $625 million payment from Verizon will initially fund the creation of the joint venture. Kirkby said any remaining proceeds after funding requirements will be directed toward reducing BT’s debt, supporting the company’s ongoing efforts to strengthen its balance sheet while maintaining investment in next-generation fiber and mobile infrastructure.

The market reacted positively to the announcement, with BT shares rising around 1% in early London trading.

The transaction highlights the changing economics of the global telecommunications sector. Enterprise customers are increasingly purchasing integrated networking, cybersecurity, artificial intelligence-enabled network management, and cloud services rather than traditional voice and data connectivity alone.

That evolution has put pressure on telecom operators to achieve greater scale, invest heavily in software-defined networking, and partner with cloud providers while protecting margins.

The joint venture reduces exposure to BT’s business, which had become an earnings drag, while preserving participation in future growth through equal ownership. The agreement also significantly expands Verizon’s international enterprise footprint without requiring a full acquisition, giving it broader global reach as multinational corporations continue to increase demand for secure digital infrastructure.

Additionally, the deal places both companies in a position to compete more effectively against global enterprise networking providers, hyperscale cloud operators and managed service firms as demand for cross-border connectivity, cybersecurity and AI-enabled enterprise services continues to accelerate.

Markets Find Temporary Relief as U.S.-Iran Ceasefire Eases Energy Fears, Though Inflation and AI Valuation Concerns Linger

0

Global markets found a measure of relief on Monday after the United States and Iran agreed to halt recent hostilities and renew diplomatic talks, helping to calm oil prices that had spiked earlier in the day amid fresh exchanges of strikes between the two sides.

The return to negotiations followed several days of tit-for-tat military actions sparked by an Iranian projectile striking a cargo vessel in the Strait of Hormuz last week. Both nations had accused each other of violating an interim ceasefire, raising fears of prolonged disruption to one of the world’s most critical energy arteries. The latest pause in fighting offered investors a window to reassess positions, leading to modest gains across equities even as underlying economic concerns persisted.

Europe’s STOXX 600 index rose 0.1% in morning trading, while futures for the U.S. S&P 500 climbed 0.7%. Asian markets also pared earlier losses, with South Korea’s KOSPI down 0.2% and Japan’s Nikkei up 0.15%. Oil initially climbed on Monday following weekend strikes but then reversed, trading near its lowest level since the conflict began. Brent crude was little changed at $72.20 a barrel, marking a 22% decline for the month.

“The market can take some relief in the lower oil prices and its impact on the global economy,” said Mohit Kumar, chief European economist at Jefferies. “Lower oil prices should lead to a diversification trade and growth-sensitive sectors which have suffered in the last few months should outperform.”

This de-escalation provides a timely counterbalance to persistent worries about stretched valuations in AI-related stocks and rising expectations for tighter monetary policy in the United States. Easing energy costs could help moderate some inflationary pressures, but recent data have shown inflation jumping in the U.S. and elsewhere, reinforcing bets that the Federal Reserve may need to raise rates to contain price growth.

Rising odds of a rate hike have supported the dollar, with the dollar index trading at 101.25, just below the one-year high it touched last week. The Japanese yen slipped slightly to 161.80 per U.S. dollar, as fears of potential intervention from Tokyo prevented it from breaking through its lowest level in 40 years.

Investors are now pricing in at least one Federal Reserve rate hike this year, a sharp reversal from expectations of two cuts before the conflict intensified. Bank of America strategists have adopted an even more hawkish stance, forecasting three hikes, partly reflecting resilient U.S. jobs growth.

The stronger dollar has weighed on gold, which fell 0.6% to $4,061 per ounce. The yellow metal is on track for a 13% decline in the second quarter — its biggest quarterly drop since 2013.

Tech Valuations and Sector Rotation Dynamics

Investor unease over AI-related valuations has also lingered. Futures for the tech-heavy Nasdaq rose 1% on Monday, putting the U.S. index on track for a rebound after slumping more than 4% last week. The Bank for International Settlements has cautioned about the durability of the current AI investment surge, noting that supply bottlenecks and intense competition could lead to overinvestment reminiscent of previous boom-and-bust cycles.

“For this reason, traders have gravitated toward the defensive and cyclically oriented areas of the equity space in recent weeks,” said Jose Torres, senior economist at Interactive Brokers.

The ceasefire news encouraged some rotation back into growth-sensitive sectors that had been under pressure from higher energy costs and tighter financial conditions. This shift could gain further traction if lower oil prices translate into reduced input costs and improved consumer confidence.

U.S. Treasury yields were little changed on Monday as investors looked ahead to key jobs data later in the week while monitoring the fragile pause in U.S.-Iran hostilities. The benchmark 10-year Treasury yield rose less than a basis point to 4.376%, while the 2-year yield rose just over 1 basis point to 4.102%.

The 30-year bond yield declined less than one basis point to 4.861%. The bond market will be closed on Friday ahead of Independence Day celebrations. Investors will closely parse May’s JOLTS job openings data on Tuesday and the June nonfarm payrolls report on Thursday to gauge the health of the U.S. economy and refine their expectations for Federal Reserve policy.

A Fragile Diplomatic Window

The U.S. and Iran agreed to pause hostilities and allow commercial vessels to freely pass through the Strait of Hormuz following military clashes over the weekend that had threatened to derail negotiations aimed at ending the conflict.

“Technical talks are slated to continue on all areas of the MOU,” a U.S. official told CNBC on Sunday. “Both sides will stand down for now, and vessels can move freely.”

A sustained return to diplomacy could ease energy market tensions and reduce inflationary pressures globally. However, the situation remains fluid, with both sides quick to accuse the other of violations in recent days. Markets will be watching closely to see whether the latest pause holds and leads to a more comprehensive agreement.

For now, the de-escalation has provided investors with a reason for cautious optimism, allowing some rotation back into growth-sensitive areas that had been under pressure. Yet underlying concerns about inflation, monetary policy, and AI valuations suggest the path ahead remains uncertain.