Posts

AI News Flash: This is a one-time task, just produce the Traditional Chinese summary directly. British luxury phone brand Vertu has launched a foldable phone called Alphafold, starting at $6,880, built with real calfskin leather and titanium, targeting enterprise executives. The selling point isn’t specs — it’s status. Vertu says its core appeal isn’t hardware, but the built-in Hermes Agent — an AI agent built on top of the open-source Hermes…

AI News Brief: In late May, Index Ventures co-founder Neil Rimer said in an interview in Athens that he strongly feels the wealth generated by AI will undergo some form of ‘redistribution’—whether voluntary or involuntary—and he hopes it will be voluntary, believing tech leaders can play a leading role. Rimer has stepped back from day-to-day investing since 2021 and now lives mostly in Athens. Despite his low profile, Ind…

AI News Brief: Morgan Stanley Nasdaq tech stocks plunged on Friday, triggered by a Chinese startup releasing a low-cost AI model whose performance is seen as capable of challenging top U.S. systems, raising doubts about America’s dominance in the AI race. The report focuses on Moonshot AI’s newly released Kimi K3 model, highlighting how Chinese developers have been rapidly closing the technical gap. The core investor concern is that Chinese teams are not only narrowing the technology gap with their U.S. counterparts but can also deliver AI services at significantly lower prices, putting pressure on U.S. AI companies’ business models and valuation logic, dragging down stocks like Nvidia. The original summary did not provide technical details such as Kimi K3’s specific parameter scale, training cost, or benchmark scores. See the original link for details.

AI news brief: Amid escalating South China Sea sovereignty disputes in 2026, Chinese state media has sparked strong backlash and boycott calls in the Philippines over an animated online video accused of carrying racist undertones. Manila is pushing for the takedown of the Facebook post criticized as ‘dehumanizing’ and pressuring the platform. The controversy comes on the tenth anniversary of the South China Sea arbitration ruling — in 2014 the Permanent Court of Arbitration rejected China’s sovereignty claims over most of the South China Sea, yet a decade on, territorial tensions between the Philippines and China in the region continue to simmer with no signs of easing.

AI News Brief: Ishihara Sangyo will invest in its Mie Prefecture factory in Japan to double the production capacity of raw materials used in AI server components. The company holds a high market share in titanium dioxide materials for ceramic capacitors, which are key raw materials for manufacturing multilayer ceramic capacitors (MLCC). MLCCs are widely used in AI server motherboards and power modules, and with AI server demand growing rapidly, the related component supply chain is facing pressure to expand. The original summary only mentions…

AI News Brief: Reporting from Tokyo in Taiwan time, Japanese chipmaker Rapidus and US software company Cadence announced on Friday that they will collaborate to offer customers chip design tools with agentic AI capabilities. This partnership is part of Rapidus’s strategy as a foundry to improve customer service convenience as it pushes toward mass production. Rapidus President Atsuyoshi Koike said AI agents will play a central role in accelerating the chip design process. The original summary did not provide further details on the specific technical architecture, product names, rollout timeline, or commercial terms of the collaboration; please refer to the original link for full details.

AI News Brief: Moonshot AI announced its Kimi K3 model performs on par with — and on some benchmarks surpasses — the latest offerings from Anthropic and OpenAI, while significantly cutting costs. The launch, timed with Shanghai’s World AI Conference, highlights how fast China’s AI industry is closing the gap with — and on certain metrics overtaking — US frontier labs, especially on cost efficiency.

AI News Brief: OpenAI CFO Sarah Friar proposes a practical AI ROI scorecard to measure real benefits of enterprise AI adoption, replacing gut-feel judgments with four quantifiable metrics: useful workload, cost per successful task, reliability, and compute resource ROI.

Google’s AI Mode now links to Instacart Canva and YouTube letting users complete real tasks across apps while stepping up competition with OpenAI and Anthropic.

AI News Brief: General Compute, a startup focused on AI inference cloud computing, just secured a $400M loan from tech investment firm Upper90 — likely the first financing deal using ‘inference-specific chips’ as collateral. These chips are designed to run trained AI models quickly and efficiently, costing far less than the high-end chips used for training. The deal reflects the market’s response to concerns about inflated AI tool and token pricing, shifting toward infrastructure that can run open-source models at lower cost rather than the latest frontier lab LLMs. General Compute was founded by CEO Finn Puklowski and closed a $15M seed round in May, building an inference-focused neocloud (distinct from general-purpose hyperscalers like AWS and Azure) around Intel-backed chipmaker SambaNova. Its SN50 chip is purpose-built for inference — power-efficient and requiring no expensive water cooling — enabling faster deployment across more diverse data centers than GPUs, with claims of 16x faster inference than GPU clouds. The challenge is how startups can secure enough chips. Upper90 co-founder and CEO Billy Libby, a former Goldman Sachs quant, provided a GPU procurement loan to energy-focused data center startup Crusoe in 2021, widely considered the first loan using advanced chip value as collateral; now that the GPU market has matured or possibly over-procured, Upper90 is betting on the next growth wave in inference. Libby points out that open-source models are rising in importance — services like OpenRouter and Fireworks have recently closed rounds at high valuations, new models like Kimi K3 match Anthropic and OpenAI’s latest versions on coding benchmarks, and chipmakers like Groq and Cerebras are drawing attention from acquirers and public markets.