Um estudo holandês investigou se cães ou gatos trazem mais benefícios para o humor dos donos. A interação com ambos os animais melhora o humor a curto prazo, mas não reduz o estresse. Gatos, em particular, podem até piorar a situação, segundo a pesquisa.
Amazon, Google, and Microsoft admit AI startup Anthropic leads in "frontier" AI models, especially for coding. However, executives state they are not concerned, citing specialized foundations and distinct development timelines to close the gap.
Palantir CEO Alex Karp expressed frustration with leading AI labs like OpenAI and Anthropic, accusing them of "tokenmaxxing" instead of focusing on business needs. He noted corporate clients are unhappy with the current AI development trajectory, emphasizing the need for practical deployment over raw model development.
OpenAI has appointed Dean Ball, a former senior policy advisor on AI in the Trump administration, to lead its new Strategic Futures unit. Ball will focus on shaping OpenAI's frontier AI policy and internal governance, citing the company's key breakthroughs and talent as reasons for joining.
Researchers are testing a geoengineering method in Cambridge Bay, Canada, to refreeze Arctic sea ice. The Real Ice project pumps ocean water onto the ice surface, thickening it and reflecting more sunlight. While controversial, it's a research effort to explore potential solutions to climate change.
Z.ai launched GLM-5.2, a powerful open-source AI model trained on Huawei Ascend chips, showing strong performance on benchmarks like FrontierSWE and SWE-bench Pro. The model boasts a 1 million-token context window and an MIT license, with competitive API pricing and potential for local deployment.
TikTok Shop rivoluziona l'e-commerce in Italia, superando i 21mila venditori attivi e attirando un pubblico transgenerazionale. La piattaforma punta su shoppable video e live shopping, diventando un trampolino per le imprese e influenzando tendenze culturali e linguistiche, pur mantenendo un focus sulla tutela dei minori.
Amazon's SVP Peter DeSantis stated the company aims to rival OpenAI and Anthropic in AI within a year, while founder Jeff Bezos refuted claims that AI will cause mass layoffs, predicting instead a new era of wealth and job creation.
AI startup Anthropic joins Frontier's carbon removal collective with a $915 million pledge, nearly doubling Frontier's total funding to $1.8 billion. This marks the first AI-only company to join the group, which vets and contracts carbon removal projects.
The White House imposed export controls on Anthropic's advanced AI models, Fable 5 and Mythos 5, causing chaos for users and casting doubt on the company's future. Conflicting narratives and internal political struggles within the administration have obscured the reasons and responsibility for the decision, highlighting the volatile nature of AI regulation.
China released a white paper advocating for "more just and equitable" global governance, particularly in emerging areas like AI and outer space. Top diplomat Wang Yi stated the current international system is at a "critical crossroads" due to multiple crises, emphasizing China's support for changes aligned with current realities rather than rebuilding the existing order.
CEOs of major AI companies, including OpenAI, Anthropic, and Google DeepMind, are attending the G7 summit in France to discuss frontier AI risks, infrastructure, and child protection online. Their presence signals a shift in geopolitical influence and an opportunity to shape global policy before binding rules are established.
微軟正式推出代理式AI工作系統「Copilot Cowork」，能端到端完成複雜、跨工具協作任務。該系統採用雲端託管、內建Work IQ、企業級安全、多模型設計及低成本等五大特色，並預計推出新模型Cowork 1。
Uno studio dell'AI4I rivela che i modelli AI avanzati, pur bloccando la maggior parte degli attacchi, restano vulnerabili a tecniche adattive. I metodi tradizionali sono inefficaci, ma gli attacchi che apprendono dalle risposte del modello continuano a ottenere risultati dannosi.
Revisão sistemática com meta-análise da Unesp concluiu que creatina não reduz marcadores inflamatórios em humanos de forma consistente. Benefícios pontuais foram observados apenas em atletas submetidos a exercícios extenuantes.
Rio de Janeiro's IT agency IplanRIO launched Rio 3.5, a 397B parameter AI model, claiming it surpassed leading models. Nex-AGI, an AI alliance, revealed Rio 3.5 is largely a merge of their Nex N2 Pro model and Alibaba's Qwen 3.5, sparking an attribution dispute.
Anthropic's temporary shutdown of its powerful AI models at Washington's request has sparked international concerns about US dominance and control over critical AI technologies. Governments worldwide are now prioritizing domestic AI development to ensure national security and technological sovereignty.
L'associazione dei camionisti della Bolivia denuncia la morte di almeno tre autisti dall'inizio delle proteste e dei blocchi stradali avviati da oltre un mese per chiedere le dimissioni del presidente Rodrigo Paz. Oltre 600 camion sono bloccati e gli autisti si sentono presi in ostaggio.
Microsoft CEO Satya Nadella argues that the true AI advantage lies not in selecting the best frontier model, but in building a proprietary 'learning loop' around it. This system, trained on a company's unique data and judgment, creates defensible intellectual property, contrasting with the subscription-based model of frontier models alone. Nadella warns against ceding all value to a few AI models, drawing parallels to the economic consequences of early globalization.
Microsoft's '2026 Work Trend Index' reveals that as AI agents become more prevalent, 'Frontier Professionals' (16% of respondents) who redesign work and build repeatable AI usage patterns achieve significantly higher performance. The report emphasizes that human judgment and leadership become more critical as AI handles more execution, but organizational systems and leadership alignment are lagging, creating a 'paradox of transition'.
Microsoft CEO Satya Nadella posted an essay on X emphasizing the importance of building a "frontier ecosystem" for AI, warning against the concentration of economic value in a few AI models. This follows a US government directive that halted access to Anthropic's AI models for all users due to export control concerns.
Meta's Chief AI Officer Alexandr Wang revealed that the company's open-source AI strategy is no longer suitable for frontier models like Muse Spark, which was kept proprietary due to bio and safety risks. Rival labs face similar issues as models scale. Meta is also exploring subscription models for its platforms and AI chatbot to diversify revenue beyond advertising.
Meta's Chief AI Officer Alexandr Wang admits the company's open-source AI strategy is outdated for frontier models like Muse Spark, which was kept proprietary due to bio and safety risks. Meta is exploring subscription models for its platforms and AI chatbot to diversify revenue beyond advertising.
Meta's Chief AI Officer Alexandr Wang stated their open-source AI strategy is unsuitable for frontier models like Muse Spark due to bio and safety risks. The company is exploring subscription models for its platforms and AI chatbot to diversify revenue beyond advertising.
