AI News Roundup – August 01, 2025

Google rolls out Gemini Deep Think AI, a reasoning model that tests multiple ideas in parallel

Google has rolled out Gemini 2.5 Deep Think, an advanced multi-agent AI system that spawns multiple AI agents to work on a problem simultaneously and explore different solutions in parallel

techcrunch Aug 01, 2025

Data Labeling Is the Hot New Thing in AI

‘Data labeling,’ the process of using human experts to fine-tune AI models, is becoming increasingly critical for developing safe and effective agentic AI, as shown by Meta’s massive investment in Scale AI

ieee Aug 01, 2025

A.I. Researchers Are Negotiating $250 Million Pay Packages. Just Like N.B.A. Stars

The intense competition for top AI talent has led to massive pay packages reaching up to $250 million, with salaries now comparable to those of professional sports stars

the new York times Jul 31, 2025

The AI Company Capitalizing on Our Obsession With Excel

A new startup has launched Shortcut, an application that looks and feels like Microsoft Excel but features a built-in agentic AI for creating complex financial models from natural language

the wall street journal Jul 31, 2025

Personal Superintelligence

Mark Zuckerberg has shared his vision for ‘personal superintelligence,’ where advanced AI acts as a personal coach to help individuals achieve their unique goals and aspirations

meta Jul 30, 2025

What’s new in NotebookLM: Video Overviews and an upgraded Studio

Google’s AI research assistant, NotebookLM, has been updated with ‘Video Overviews’ to auto-create narrated presentations from documents and an upgraded Studio panel for storing multiple outputs

google Jul 29, 2025

Introducing study mode

OpenAI has launched a new ‘study mode’ in ChatGPT, designed to foster deeper learning through interactive, step-by-step guidance instead of providing immediate answers

openai Jul 29, 2025

Microsoft Releases List of Jobs Most and Least Likely to Be Replaced by AI

A new Microsoft paper suggests that jobs centered on information and writing are most vulnerable to AI automation, while manual labor roles are considered the safest from replacement

futurism Jul 29, 2025

China’s AI race heats up as Shanghai launches massive subsidy scheme

Shanghai is intensifying China’s AI race with a new $139 million subsidy program designed to lower operating and computing costs for local AI startups and businesses

south china morning post Jul 29, 2025

China urges global consensus on balancing AI development, security

China is calling for a global consensus on balancing AI development and security, announcing a new international cooperation body and offering to share its advancements with developing countries

techxplore Jul 26, 2025

Get the latest updates on AI advancements, research, and applications delivered straight to your inbox