-
DeepSeek finds a new way to improve AI memory using visual tokens. There is a call for scientists to create a positive vision for AI. Big Tech (Meta, Google, Microsoft) is tripling down on AI spending, while Nvidia became the first $5 trillion company. Google upgrades NotebookLM with custom chat goals and a 1M token…
-
AI Weekly (October 24, 2025): OpenAI introduces ChatGPT Atlas, a new browser that directly challenges Google Chrome. A massive AWS outage is being linked to recent AI-related layoffs at Amazon. A new report details the hidden data trail left by agentic AI, while a study warns that AI models can get “brain rot” from low-quality…
-
AI Weekly (October 17, 2025): Wall Street is sounding the alarm on a potential AI market bubble. Google’s NotebookLM gets a visual upgrade with the Nano Banana image model. AI chatbots are rapidly replacing call center workers in India. Anthropic releases Haiku 4.5, a free and powerful new small language model. A new essay argues…
-
AI Weekly (October 10, 2025): A new report finds 26% of jobs could be radically transformed by AI. Google introduces its Gemini Enterprise platform to help businesses build custom AI agents. Consulting firm Deloitte was forced to repay the Australian government for an AI-generated report. OpenAI CEO Sam Altman warns the AI industry is due…
-
Artificial Intelligence (AI) is transforming Latin America and the Caribbean. The 2025 Latin American Artificial Intelligence Index (ILIA) assesses 19 countries across key areas like infrastructure, talent, and governance. It reveals the region’s awakening to AI but warns that without investment and execution, Latin America risks consuming, rather than creating, its own solutions.
-
AI Weekly (October 3, 2025) : Anthropic releases Claude Sonnet 4.5, its best coding model yet. California passes the first comprehensive AI safety law in the US. OpenAI unveils Sora 2, a major leap in AI video and audio generation. Reuters launches AI-generated earnings summaries. OpenAI introduces a new metric to evaluate AI performance on…
-
Q3 2025 was a turning point for AI. Agents moved from hype to everyday tools, governments scrambled for control, and soaring energy demands revealed the hidden costs of progress. From finance to filmmaking, AI seeped into daily life—while unease over jobs and trust lingered. The quarter showed AI is no longer optional, but unavoidable.
-
AI Weekly (September 26, 2025) Government AI initiatives will fail without workforce upskilling. Figma makes its design tools more accessible to AI agents. The creators of NotebookLM launch Huxe, a new audio-first research app. Alibaba challenges top US models with its new multimodal AI, Qwen3-Omni. Advanced AI can now pass all three levels of the…
-
AI Weekly (September 19, 2025) Top AI scientist Song-Chun Zhu’s move to China highlights the geopolitical race for talent. Google releases VaultGemma, its first privacy-preserving LLM. An ethics brief warns of growing ties between Silicon Valley and the military. A study shows most people use ChatGPT for practical, everyday tasks. Google releases a 10-point policy…
-
AI Weekly (September 12, 2025): AI pioneer Geoffrey Hinton warns of massive AI-driven unemployment. Google’s NotebookLM adds new AI-powered study tools for students. The use of AI in filmmaking is ramping up, sparking debate in Hollywood. A new system called RSL aims to make AI companies pay for web content. OpenAI is reportedly using AI…
AI Talks 