AI Brief

Samsung Group’s trading arm, including Samsung C&T, is partnering with Japan’s major power company Erex to build a biomass power plant in Japan, aiming to provide stable renewable energy supply for the grid amid rapidly growing data center electricity demand driven by AI. The plant will be operated by a special purpose company (SPC), with biomass fuel primarily imported from Southeast Asia. This collaboration reflects how Japanese companies are actively pursuing renewable energy solutions capable of stable baseload power rather than relying solely on intermittent solar or wind, targeting stable supply before FY2029 to meet continued data center demand expansion. The original article does not specify plant capacity, investment amount, or completion date; see the source link for full details.

AI News Brief: Global M&A deals reached $2.84 trillion from January to June, setting a new half-year record and surpassing the previous high set in the first half of 2021. This M&A wave is primarily driven by large US tech companies actively acquiring stakes in AI and adjacent fields (such as energy) to seize leadership, reflecting how the AI race is accelerating corporate consolidation of compute, energy supply, and technology footprints through M&A. See the original article for full details.

AI news brief: Hyperscale Data (formerly Ault Alliance, renamed and repositioned around AI infrastructure in 2024 while keeping its Bitcoin mining operations) announced on Thursday that it sold roughly 100 Bitcoin and secured a Bitcoin-collateralized credit line, with proceeds from both going toward building out its Michigan AI data center. Funds from the Bitcoin sale will go to construction work and procurement of critical infrastructure and long-lead equipment…

AI news brief: A story about Japanese-rule-era transitional dementia patients meeting at a flea market in Beijing’s Forbidden City and starting a collaboration. The author uses the stacked sessions and pull requests features in the GitHub Copilot app to modernize a legacy codebase. See the original link for details. …

AI News Brief: Gemini Robotics ER 2 is an upgraded robot-specific model from Google DeepMind, focused on enhancing robots’ ‘reasoning’ and ‘collaboration’ capabilities in real-world environments. The core breakthroughs span three areas: first, significantly improved video understanding that lets robots more accurately parse dynamic scenes and object interactions rather than relying solely on static image judgment; second, tool orchestration capability, enabling robots to autonomously plan and chain multiple tools or subtasks to complete complex commands without needing every step manually broken down; third, multi-robot collaboration, letting different robots coordinate and divide work to jointly execute the same task — especially critical for scenarios like warehouse logistics and production lines that need multiple robots operating in sync. Overall, this represents a step forward from a single robot executing a single command toward systems that understand complex situations, autonomously plan steps, and cooperate with other robots. The original article summary itself is relatively broad and doesn’t provide specific performance numbers, benchmark scores, or technical architecture details; see the original link for the full content.

AI News Brief: This news piece is very short, only an introductory paragraph with limited substantive details. The Trump administration once proposed banning Chinese open-weight AI models from use in the US, but before China’s tech industry and officials could take a stance, Silicon Valley had already stepped forward to express opposition. The report points out that the Trump administration and US tech companies hold different views on how to respond to the rise of Chinese AI models: the government favors restrictive measures like bans to block them, but Silicon Valley players including Nvidia believe that amid growing AI bubble concerns and market caution over growth momentum, the US tech ecosystem itself still depends on open-weight models to sustain innovation and expansion. A hasty ban on Chinese open-weight models could end up hurting the development foundation and competitiveness of America’s own AI industry. The original summary only provides event background and an overview of opposing positions, without further detailing the specific proposal contents, timeline, or the full arguments from Silicon Valley’s opposition. Please refer to the original link for details.

AI news brief: Zhongji Innolight (中際旭創) is a Chinese optical communications component supplier that mainly produces AI-related key equipment. The company debuted on the Hong Kong Stock Exchange on July 30, with shares falling on the first day. The IPO raised HK$53.4 billion (about US$6.8 billion), making it Asia’s second-largest listing so far this year. The bell ceremony was attended by Chairman and President Liu Sheng along with other executives. The weak first-day stock performance reflects market concerns about…

AI News Brief: Japan has posted a trade deficit for two consecutive months, while the Philippines’ monthly deficit continues to widen—Thursday’s data from both governments lays bare a divergence in Asia’s trade landscape. Rising crude oil prices are squeezing trade balances and currencies across multiple Asian economies, and the numbers show a clear split: some economies are riding tech export growth hard enough to offset the energy cost shock, while others can’t fully absorb it, and their trade deficits are widening as a result. Japan and the Philippines fall squarely in the latter camp—both saw their trade accounts deteriorate in June, reflecting the dual pressure on trade balances and local currencies from rising oil prices layered on top of uneven tech export momentum. The original summary does not provide specific deficit figures, year-over-year changes, or which tech export items are driving other Asian economies’ performance, so details will need to be verified against the original source.

AI News Brief: Teleperformance, a France-based business process outsourcing (BPO) group, revealed to Nikkei Asia that the company plans to open its first site in Vietnam before the end of this year. The move primarily targets demand related to AI data services, as AI technology is reshaping how the call center industry operates, prompting the company to adjust its strategic positioning in response to structural market shifts. The original summary only outlines high-level plans without further details on site scale, investment amount, staffing, or specific AI data service content — see the original link for full details.

AI News Brief: SoftBank Corp. and NTT are partnering with dozens of Japanese companies and research institutions to drive a nationwide AI and data sharing initiative, aiming to build a foundational technology platform for cross-industry data integration. This move is seen as part of Japan’s strategy to catch up with the US and China in AI. Since Japan lags behind the US and China in AI model development, the government and industry hope to leverage the scattered data resources held by various companies and institutions to build AI infrastructure with scale and differentiation…