Yinhua Hang Seng HK Equities China Technology ETF

The article draws a parallel between the 1970s fuel shortage and the current memory shortage, urging software engineers to prioritize efficiency. As memory prices rise, the author argues that it's time to reconsider the bloated nature of modern software. The piece calls for a shift in mindset towards creating more compact and efficient applications, similar to those from earlier computing eras.

The Register

Memory is running out, and so are excuses for software bloat

The article by Jacob Zhao discusses the evolution of artificial intelligence from statistical learning to structured reasoning, emphasizing the role of reinforcement learning. It highlights the emergence of DeepSeek-R1, marking a paradigm shift in reinforcement learning, and outlines its architecture and applications. The article details the stages of reinforcement learning, including policy exploration, preference feedback, reward modeling, and policy optimization, and introduces new optimization methods like GRPO and DPO, showcasing advancements in AI decision-making capabilities.

- Artificial intelligence is transitioning from pattern fitting to structured reasoning, highlighted by the emergence of DeepSeek-R1 in reinforcement learning.  
- Reinforcement learning is crucial for decision-making improvement across various industries, with applications in areas like gaming, robotics, digital reasoning, scientific discovery, and finance.  
- Advanced frameworks such as GRPO and DPO enhance policy optimization and learning processes, showing significant technological progress and the potential for future development in decentralized systems.

CoinLive

C3.AI

From Computing Power to Intelligence: A Decentralized AI Investment Map Driven by Reinforcement Learning

A newly developed technology called 'DSC' now allows users to create 3D spatial content using their mobile devices. This platform supports data-driven operations by incorporating artificial intelligence and extended reality. It is designed to aid enterprise planning, operations, and management by going beyond traditional digital twins. The platform will be showcased at a booth displaying operational use cases based on extended reality.

Unusual Whales

DEEP.FINE to debut AI and XR platform 'DSC' at CES 2026, enabling 3D creation on mobile, supporting data-driven ops.

Paul Keary, CEO and cofounder of Teneo, says CEOs are taking a longer-term view of AI as a catalyst for business transformation. He tells Romaine Bostick and Katie Greifeld on "The Close" that after a year focused on AI capital spending, the next phase will center on how enterprises transform using AI.Bloomberg videos, provided by MT Newswires

- Paul Keary, CEO of Teneo, emphasizes that CEOs are increasingly viewing AI as a long-term catalyst for business transformation.  
- In an interview, he highlights that after a year dedicated to spending on AI, the next focus will be on how enterprises leverage AI for transformation.  
- This shift marks a significant evolution in the approach toward AI within organizations.

marketscreener

Teneo CEO: AI The Largest Spend Item for 2026

Organisations are missing out on 55% of potential productivity gains from AI adoption due to a disconnect between AI use and workforce readiness, according to the EY 2025 Work Reimagined Survey. The survey highlights the need for better AI training and integration, as well as addressing workplace anxieties. Despite widespread AI use, only a small percentage of employees use it to enhance workflow. The report also notes a rise in talent satisfaction and a drop in resignation rates, but stresses the importance of leadership and career development to retain AI-skilled talent.

- The EY 2025 Work Reimagined Survey reveals that organizations miss 55% of potential productivity gains from AI due to a gap between adoption and workforce readiness.  
- While 90% of employees use AI, only 5% receive adequate training to leverage it effectively, and concerns over skill erosion and misinformation persist.  
- To maximize productivity and talent retention, organizations must invest in AI training and create supportive environments for effective integration.

Singapore Business Review

Organisations miss 55% productivity gains from AI adoption

Artificial Intelligence (AI) is transforming industries, economies, and our everyday lives at an unprecedented pace. As a key driver of

Invest $10,000 in these AI stocks for long-term growth

Sabrina Kopecki, an AI Technical Specialist at IBM, emphasized the need for a unified security and governance framework for Agentic AI, which actively performs tasks. She highlighted the risks of Shadow AI, unofficial AI solutions that bypass formal IT channels, creating vulnerabilities like data leakage and compliance issues. Kopecki proposed a continuous oversight cycle of Discover, Assess, Govern, Secure, and Audit to ensure AI systems remain stable, accountable, and safe. She illustrated this with use cases in healthcare and the public sector, advocating for predictable AI operations that deliver value without future crises.

StartupHub

Agentic AI Demands Unified Oversight: Taming Shadow AI with Zero Trust

An analysis by Egan-Jones in New York delves into the potential contrasting impacts of artificial intelligence from conventional beliefs. Despite the common narrative of AI leading to job elimination and overall prosperity, Egan-Jones provides an alternative perspective on the probable distribution of benefits from AI integration. The commentary highlights that AI advancements are commonly suggested to outperform humans in various realms, including legal assessments and medical diagnostics.

Egan-Jones analyzes AI's impact on economics, predicts different distribution of gains. AI surpasses human performance.

NVDA

MSFT

GOOG

GOOGL

Daniel Newman, CEO of Futurum, argues that the AI market is in a 'supercycle' rather than a bubble, driven by real revenues from companies like Nvidia, Google, and Microsoft. He distinguishes between AI hype and genuine revenue drivers, noting that some companies may inflate valuations by overusing 'AI' in earnings calls. However, key infrastructure firms are generating substantial revenues, indicating a long-term technological growth cycle.

NVIDIA

Microsoft

Alphabet - C

Alphabet

Futurum CEO Daniel Newman believes AI market is in a 'supercycle,' not a bubble. Core companies like Nvidia, Google, and Microsoft are driving real revenues.

Daniel Newman, CEO of Futurum, argues that the current AI market is not a bubble but a multi-decade 'supercycle.' He acknowledges risks like overvaluation and overspending but emphasizes that core infrastructure companies like Nvidia, Google, and Microsoft are generating real revenues. Newman predicts 2026 as a pivotal year for AI revenue proof and warns of potential risks if demand decreases. His views contribute to the ongoing debate about the AI market's stability.

- In a recent interview, Daniel Newman, CEO of Futurum, stated that the current AI market is not in a bubble but part of a "multi-decade technological supercycle."  
- He acknowledged concerns about inflated valuations and spending by some companies but highlighted that core infrastructure firms like Nvidia, Google, and Microsoft are driving substantial real revenues.  
- Newman indicated that 2026 would be a pivotal year for demonstrating AI monetization in enterprises, despite potential supply-side constraints posing risks to sustained demand.

benzinga_article

Futurum CEO Daniel Newman Calls Current AI Market Multi-Decade 'Supercycle,' Not A Bubble— But Sudden Glut 'Could Become A Risk'

According to a recent prediction by Gartner, vector databases are anticipated to be incorporated into about 30% of enterprise applications by the year 2026. Concurrently, Milvus, an open-source vector database developed by Zilliz, has achieved a significant milestone by surpassing 10,000 production deployments. This update was unveiled in Redwood City, California on December 18, 2025.

Milvus hits 40,000 GitHub stars, with 10,000+ productions, strengthening its position as a leader in vector databases. Gartner predicts 30% enterprise usage by 2026.

OpenAI

The AI landscape is rapidly evolving with breakthroughs in model efficiency and geopolitical competition for computational supremacy. Google, NVIDIA, OpenAI, Zoom, and Meta are making significant strides in democratizing AI, enhancing accessibility, and optimizing performance. Political tensions arise over AI infrastructure demands, with calls for balanced policy. Major investments in compute resources underscore the competitive race for AI dominance. The strategic orchestration of AI capabilities is proving crucial, with companies leveraging diverse models for optimal performance.

- The AI landscape is rapidly evolving with significant advancements in model efficiency and innovative architectures, driven by competition among industry leaders.  
- Google introduced Gemini 3 Flash, aiming to democratize powerful AI with high performance at a low cost, while NVIDIA's Nemotron 3 models focus on open-source customization.  
- Political debates, notably a call for a moratorium on data centers by Senator Bernie Sanders, highlight tensions between technology advancement and governance, amidst intense competition for AI compute resources.