新科技/风口

Microsoft AI CEO Mustafa Suleyman poaches three Google DeepMind former colleagues, including two who built NotebookLM's Audio Overviews and worked on Astra (t.co)
2025-2-5 Google Microsoft AI

Permalink to Forrester: Finding AI talent is challenging (sdtimes.com)
2025-2-5 AI

Permalink to DeepSeek R1 is now available on Azure AI Foundry (sdtimes.com)
2025-2-5 Azure AI

Google Removes Pledge To Not Use AI For Weapons From Website (tech.slashdot.org)
2025-2-5 Google AI

Sources: the head of the US AI Safety Institute, Elizabeth Kelly, is set to leave by the end of the week; Kelly was appointed to the position in February 2024 (www.bloomberg.com)
2025-2-5 AI

AI-Generated Slop Is Already In Your Public Library (news.slashdot.org)
2025-2-5 AI

The EC releases its guidelines on prohibited AI practices under the EU's AI Act, such as using AI to track employees' emotions or manipulate website users (www.reuters.com)
2025-2-5 AI

Robotics startup Figure is exiting its Feb. 2024 OpenAI deal, which aimed to develop AI models for robots, to focus on in-house AI after a “major breakthrough” (techcrunch.com)
2025-2-5 Startup AI Robot

Sources: Meta is merging the teams behind Facebook and Messenger into one unit and shuffling its generative AI group, as it prepares for layoffs next week (www.theinformation.com)
2025-2-5 Facebook AI Layoffs

Google drops language from its AI Principles that said it would not pursue applications “likely to cause overall harm”, including weapons and surveillance (www.washingtonpost.com)
2025-2-5 Google AI

Red Hat Plans to Add AI to Fedora and GNOME (linux.slashdot.org)
2025-2-5 AI

Panasonic To Cut Costs To Support Shift Into AI (slashdot.org)
2025-2-5 AI Panasonic

Adobe adds “contract intelligence capabilities” to Acrobat's $4.99 per month AI Assistant, letting the PDF manager summarize complicated contract language (www.theverge.com)
2025-2-5 AI Adobe Intelligence

ByteDance researchers demo OmniHuman-1, an AI system that is “easily heads and shoulders above previous deepfake techniques”, if ByteDance's clips are accurate (techcrunch.com)
2025-2-5 AI

Samsung Galaxy S25 Ultra review: A polished smartphone that's so good, it just got better with AI (indianexpress.com)
2025-2-4 Samsung AI

Krutrim debuts ‘open-source' AI models to take on DeepSeek, plans to build supercomputer with Nvidia (indianexpress.com)
2025-2-4 AI Nvidia

Perplexity AI helps you tailor search experience: 5 reasons why you should try it (indianexpress.com)
2025-2-4 AI

Ola founder Bhavish Aggarwal invests $230M in his Indian AI startup Krutrim, and plans to raise $1.15B from outside investors by 2026, to build local AI (techcrunch.com)
2025-2-4 India Startup AI

DeepSeek's AI app has been banned in these places: What is the reason? (indianexpress.com)
2025-2-4 AI

AI Systems With 'Unacceptable Risk' Are Now Banned In the EU (slashdot.org)
2025-2-4 AI

British volunteer, 18, is killed within just minutes of his first mission in Ukraine, becoming UK's youngest victim of the war - as best friend describes terrifying drone encounter that claimed his life (www.dailymail.co.uk)
2025-2-4 U.K. Drone Ukraine

EU bans AI tools used for social scoring, predictive policing under landmark legislation (indianexpress.com)
2025-2-4 AI

Salesforce Cutting 1,000 Roles While Hiring Salespeople for AI (slashdot.org)
2025-2-4 AI Salesforce

CERN's Mark Thomson: AI To Revolutionize Fundamental Physics (science.slashdot.org)
2025-2-4 AI

Salt Lake City-based Jump, which offers AI tools for financial advisors and other financial services providers, raised a $20M Series A led by Battery Ventures (www.finsmes.com)
2025-2-4 AI Battery

Alphabet faces scrutiny on AI spending as Google cloud growth slows (indianexpress.com)
2025-2-4 Google AI Alphabet

Permalink to Beyond LLMs: The Power and Pitfalls of Multi-Agent AI (sdtimes.com)
2025-2-4 AI

TSA's Airport Facial-Recognition Tech Faces Audit Probe (yro.slashdot.org)
2025-2-4 Facial Recognition

Sources: Thomas Shedd, GSA's new Technology Transformation Services director and a Musk ally, told GSA workers the agency will pursue an “AI-first strategy” (www.wired.com)
2025-2-4 AI

Anthropic details Constitutional Classifiers, a system that aims to guard AI models against jailbreaks, monitoring both inputs and outputs for harmful content (t.co)
2025-2-4 AI

Previous Page Next Page