Artificial intelligence (AI) is reshaping sectors like healthcare and banking, but it poses privacy challenges, prompting G7 leaders to explore solutions. As they convene in Alberta, the focus is on developing AI systems that respect privacy without hindering innovation. Federated learning (FL) emerges as a key, often overlooked, strategy. Unlike traditional AI, which centralizes data, FL trains models locally, sharing only updates, thus mitigating data breach risks. This method has successfully been applied in Canada for cancer detection and chronic disease prediction without exposing sensitive information. As governments rush to regulate AI, FL could facilitate cross-jurisdictional collaboration on critical issues while safeguarding local data control. For democracies, endorsing FL not only signals a commitment to ethical governance but also promotes trust in AI. The G7 should prioritize FL in its discussions to foster a secure, equitable AI future, leveraging existing technologies to create a foundation for privacy and innovation.
Source link
The Case for the G7 to Adopt Federated Learning
Databricks Teams Up with Google (GOOGL) to Launch Gemini AI Models – TipRanks
Databricks has announced a partnership with Google Cloud to host its Gemini AI models on Google’s platform. This collaboration aims to enhance data analytics and machine learning capabilities, allowing organizations to leverage advanced AI tools more effectively. By integrating Gemini’s features with Google Cloud’s infrastructure, users can expect improved performance and scalability in their AI initiatives. Databricks’ advanced analytics capabilities, combined with Google Cloud’s robust technology, will facilitate the development of innovative applications and solutions. This partnership aligns with the growing demand for AI-driven insights across various industries and reinforces Databricks’ commitment to accessibility and collaboration in the AI landscape. The integration is set to benefit businesses seeking to optimize their data strategy through AI advancements, making it easier to adopt and implement these technologies. Overall, this collaboration signifies a significant step forward in the fusion of data analytics and artificial intelligence.
Source link
Gemini 2.5 Flash Lite and Pro GA Set to Launch in June, According to Leaked Information
Google is advancing its Gemini model lineup, with recent API findings indicating new releases like the Gemini 2.5 Flash Lite Preview, dated June 17. This suggests an imminent announcement, likely aimed at developers looking for quick, efficient solutions for lightweight tasks. The Gemini 2.5 Pro model, marked for June 19, hints at broader availability later that week, designed for more complex applications. The discoveries highlight ongoing internal experimentation with the Gemini 2.5 Pro, involving adjustments to system prompts and evaluation processes, often tested in environments like LM Arena, a benchmark platform. Observations confirm active testing of at least one new Google model, with details still unclear regarding whether it’s the Flash Lite or Pro version. The coming week is expected to be crucial for AI developers and researchers, as more information on these innovative releases is anticipated. Google aims to balance high performance with resource efficiency across both Pro and Lite models, catering to diverse user needs.
Source link
Apple’s Enhanced Tools for AI: A Game Changer for Developers – Computerworld
Apple’s anticipated AI coding tool, Swift Assist, has yet to launch, raising concerns about the company’s internal AI project failures. Initially teased at WWDC, developers expected collaboration with Anthropic for Xcode 26’s AI features, which hasn’t materialized yet. However, developers can still access the Anthropic API, providing flexibility in AI options. Feedback from developers indicates a positive reception to the integration of ChatGPT into Xcode 26. One developer quickly utilized the new version to create an on-device AI chatbot, leveraging ChatGPT for efficient code completion. While there were some bugs, the developer noted that the tool significantly accelerated the writing of boilerplate code, indicating that despite some hurdles, the integration could greatly benefit developers’ productivity moving forward. As the new Apple operating systems are set to launch this fall, the future of AI in Xcode remains optimistic.
Source link
Fighting Back Against Nudify Apps: Legal Action and Innovative Technology Solutions
The rise of ‘nudify’ apps, which use AI to create non-consensual explicit images, has prompted Meta to strengthen its policies against such content. The company actively removes ads and accounts promoting these services and blocks related search terms on its platforms. Meta has taken legal action against Joy Timeline HK Limited, the entity behind the CrushAI apps, for repeatedly attempting to advertise in violation of its rules. The lawsuit aims to prevent further promotions of these apps on Meta platforms. Additionally, Meta collaborates with other tech companies through the Tech Coalition’s Lantern program to share information on these violations. As tactics used by advertisers evolve, Meta employs new technologies and strategies to detect and eliminate ads promoting nudify apps. The company also supports legislation like the U.S. TAKE IT DOWN Act to combat intimate image abuse and advocates for parental controls to protect teens from downloading harmful apps.
Source link
Empowering Change: Odisha’s Young Innovator Develops ₹85 AI Tool for Protection
At just 21, Rishikesh Amit Nayak from Odisha is making a significant impact with his ₹85 AI-powered tool that detects plant diseases, aiding farmers in preventing crop failures. Growing up in an agricultural family and influenced by the 2017 agrarian crisis, he founded Kishan Know, an agri-tech platform that provides real-time crop health alerts via WhatsApp and SMS for only ₹85 per acre per month. Rishikesh’s vision extends beyond agriculture; through his AI-driven EdTech tool, Jiveesha, he empowers children with Specific Learning Disabilities from various socio-economic backgrounds. Recognized by the Government of India’s IndiaAI mission, he epitomizes grassroots innovation powered by technology. His inspiring journey showcases how one young Indian is leveraging AI to tackle substantial challenges, both in farming and education.
Source link
Transform 2D Images into 3D Printable Meshes Instantly with This Open-Source Generative AI Tool
Sourcing files for 3D printing can be either enjoyable or challenging. For those interested in creating 3D objects, mastering the necessary skills can be daunting, but tools like PartCrafter simplify the process. Developed by a collaboration from Peking University, ByteDance, and Carnegie Mellon University, PartCrafter is an open-source tool available on GitHub that generates 3D models from 2D images using AI. It leverages a database of 50,000 3D objects to accurately produce custom meshes based on user specifications regarding the number of parts.
PartCrafter not only excels in individual object creation but can also reconstruct entire room layouts, demonstrating its design versatility. Compared to similar applications, it offers significant enhancements in quality and speed. After designing your 3D object, you’ll need a suitable printer; resources are available for both beginners and experts to explore various 3D printing options or simulators.
Source link
Exploring the Influence of AI-Driven Trading Tools on Indian Retail Traders
The Indian retail trading market is rapidly transforming with the integration of AI-powered trading tools. A KPMG report from 2023 indicates that over 35% of retail traders now use AI tools, up from 12% in 2020. This growth is fueled by increased smartphone access, a surge in new investors, and a need to reduce emotional trading biases. Prominent AI tools include automated trading bots offered by platforms like BTZO and KuCoin, which function 24/7 to reduce human error. Predictive analytics and sentiment analysis tools also enhance trading accuracy significantly. AI tools simplify complex strategies, helping both novices and experienced traders. Despite these benefits, challenges like over-optimization, technical failures, and regulatory uncertainties exist. Nonetheless, AI in trading is becoming essential, with the algo-trading market expected to grow at a 12.5% CAGR. As AI solutions evolve, they will play a critical role in the future of retail trading in India.
Source link
RIAA CEO Supports Disney and Universal’s Legal Action Against Midjourney
Mitch Glazier, chairman and CEO of the RIAA, has voiced support for a lawsuit against the AI company Midjourney, which generates images from text prompts. The lawsuit, filed by Disney and Universal, alleges that Midjourney has unlawfully used numerous copyrighted works to train its image generator and allows users to create derivative images of famous characters. Glazier labeled Midjourney a “bad actor,” emphasizing the lawsuit as essential for promoting responsible innovation. He criticized AI firms that exploit human-created works for commercial purposes, stating this violates copyright law and is unfair to artists. Glazier advocates for collaborations that advance AI while honoring human creativity. As of now, Midjourney has not publicly responded to the lawsuit. In related news, producer Timbaland has launched an AI entertainment company, Stage Zero, aiming to create new artists and narratives from scratch.
Source link
Disney and Universal’s Lawsuit Against Midjourney: A Pivotal Moment for the Future of AI Art
Hollywood is taking significant action to protect intellectual property rights against generative AI technologies. Disney and Universal have jointly filed a lawsuit against Midjourney, a leading AI image generator, accusing it of copyright infringement by using iconic characters and art styles from properties like The Simpsons and Star Wars. This marks the first legal action from major studios targeting an AI company. They seek damages and an injunction to stop Midjourney’s operations, arguing it facilitates extensive plagiarism by allowing users to create derivative images of well-known characters. While some AI tools, such as Adobe’s Firefly, are trained on licensed materials, Midjourney’s model allegedly relies on unlicensed content. With the outcomes of this lawsuit potentially redefining AI’s role in creative sectors, Hollywood may also pursue in-house development of AI generators using their own content. The future of AI art generation hinges on the verdict and ongoing legal battles with AI companies.