ProPublica reveals issues with an AI system used by the Department of Government Efficiency (DOGE) in reviewing Veterans Affairs contracts. Created by Sahil Lavingia, the script incorrectly flagged contracts, such as essential internet services, for cancellation due to ambiguous instructions. Experts pointed out major flaws, including reliance on outdated AI models and failure to analyze complete contract texts. Many contracts were unjustly categorized as “munchable,” leading to potentially harmful decisions. Lavingia admitted to mistakes but emphasized that his work would undergo vetting. The push to rapidly implement a contract review mandate from the Trump administration resulted in widespread errors. Notably, contracts crucial for patient care and compliance were misidentified, reflecting deeper issues with understanding and defining core medical services. Despite these setbacks, the VA continues to advocate for AI deployment in streamlining claims processing and contract management.
Source link
Unveiling the AI Tool Behind DOGE’s Review of Veterans Affairs Contracts — ProPublica
New Tulane Study Reveals Generative AI Enhances Creativity, But Mainly for Strategic Thinkers
A study by Tulane University investigates how generative AI tools like ChatGPT can enhance employee creativity, finding that this boost depends on employees’ critical thinking skills. Conducted with over 250 employees from a tech consulting firm, the research revealed that those using AI tools produced more innovative and practical ideas, particularly when they actively engaged with the tools instead of using them passively. Lead author Shuhua Sun emphasizes the importance of “metacognitive strategies”—skills for planning, monitoring, and adjusting tasks. For organizations to maximize AI’s potential in fostering creativity, they must invest in developing these metacognitive skills, rather than just deploying AI tools. The study suggests that educational systems also need to prioritize these skills to prepare future workers for an AI-driven landscape. Effective training programs can teach employees to use AI strategies intentionally, laying a foundation for improved creativity in the workplace and beyond.
Source link
Unauthorized Access
The content appears to be an error message indicating that access to a specific webpage on Business Standard is denied. The user does not have the necessary permissions to view the article, which seems to discuss updates related to Google’s Gemini for Android, including the addition of a swipe gesture feature to launch live mode. The message includes a reference number for troubleshooting purposes but does not provide any specific details about the article or its contents. As a result, no information about the Gemini update can be summarized from this error notification.
Source link
Google Unveils Enhanced Gemini 2.5 Pro Preview: Improved Coding and Reasoning Capabilities
Google has introduced an upgraded preview of the Gemini 2.5 Pro model, set to become generally available soon. This update builds on the May release showcased at Google I/O, focusing on enterprise-scale applications. It showcases significant improvements, achieving a 24-point Elo increase on LMArena and a 35-point jump on WebDevArena, marking it as a leader in various benchmarks, including coding and academic performance.
Feedback from developers on previous iterations has been addressed, leading to enhancements in response quality and formatting. The upgraded model can be accessed via the Gemini API in Google AI Studio and Vertex AI, accompanied by new thinking budgets for cost and latency management. Additionally, Google unveiled Deep Think, which enhances reasoning capabilities, enabling the model to perform exceptionally well on the USA Mathematical Olympiad. Another innovation, Gemini Diffusion, was also announced, offering rapid output generation at 1,479 tokens/second.
Source link
OnePlus 13s Introduces Gemini Lockscreen Shortcut Ahead of Samsung Devices
Samsung has transitioned to using Gemini as the default AI assistant in its One UI 7, sidelining Bixby. The OnePlus 13s, however, has introduced a unique Gemini lockscreen shortcut that is absent from Samsung and Google devices. Launched in India, the OnePlus 13s features an array of AI capabilities such as writing assistance and message translation, along with a dedicated button for AI access. The lockscreen shortcut for Gemini doesn’t require biometric verification, unlike its Samsung counterpart. The OnePlus 13s boasts impressive specifications, including a 6.32-inch 120Hz OLED display, a Snapdragon 8 Elite processor, a 5,850mAh battery supporting 80W fast charging, and advanced camera features. Priced at INR 54,999 (~$641), it aims to compete with the Galaxy S24, which has a less powerful processor and smaller battery but superior cameras and longer software support.
Source link
OpenAI Appeals NYT Copyright Case: CEO Sam Altman Advocates for AI as Conversational Partners
OpenAI is appealing a court decision resulting from a copyright case initiated by The New York Times, which demands the company retain all ChatGPT output data indefinitely. OpenAI argues that this contradicts its commitment to user privacy. CEO Sam Altman emphasized on X (formerly Twitter) that they will resist any demand compromising user privacy, labeling it a core principle. The court directed OpenAI to preserve user-generated data, following the NYT’s request. Altman criticized the request as inappropriate and potentially setting a bad precedent. He suggested the concept of “AI privilege,” likening conversations with AI to those with lawyers or doctors. OpenAI’s COO, Brad Lightcap, echoed these sentiments, describing the NYT’s demands as overly broad and detrimental to privacy norms. OpenAI insists that the claims in the lawsuit are baseless and is committed to appealing the order to prioritize user trust and privacy.
Source link
OpenAI Report Sheds Light on 10 Threat Campaigns: Windows Malware and Fake Resumes Uncovered – TechRepublic
The OpenAI report highlights ten notable AI-driven threat campaigns, showcasing the evolving landscape of cyber threats. Among these campaigns are instances of Windows-based malware and deceptive practices, such as the use of fake resumes to manipulate hiring processes. Cybercriminals increasingly leverage AI to create more convincing phishing attacks and malware designed to evade detection. The report emphasizes the sophisticated tactics employed by these malicious actors, which pose significant risks to individuals and organizations alike. It underscores the importance of vigilance and the adoption of advanced cybersecurity measures to counter these AI-fueled threats. Additionally, the report calls for heightened awareness among businesses to recognize and mitigate these potential vulnerabilities effectively. As AI technology continues to advance, the threat landscape will likely become more complex, necessitating a proactive approach to cybersecurity.
Source link
Hey Siri, Am I Alright? How AI Tools Are Being Developed to Identify Suicidal Warning Signs
The text discusses how AI can enhance suicide risk identification and prevention. Traditional methods rely on subjective assessments, lacking predictive accuracy despite various scales. AI and machine learning offer promising solutions by analyzing large datasets to identify risk factors and patient profiles. Key goals include improving risk prediction accuracy, understanding important predictors, and modeling patient subgroups. Social media platforms, often seen as detrimental to mental health, can also provide valuable real-time data. AI tools can analyze user-generated content for signs of distress, such as expressions of hopelessness, and trigger timely interventions. Companies like Meta, Samurai Labs, and Sentinet utilize algorithms to spot suicidal ideation and provide support resources. While AI cannot replace human empathy or professional care, it serves as a crucial ally in suicide prevention, offering faster identification of warning signs and enabling early interventions.
Source link
NEET Challenges and AI Boom Drive Engineering Applications to a Five-Year High in Tamil Nadu
Engineering education is increasingly popular in Tamil Nadu, with a 14% rise in applications for counselling this year, totaling 295,134 applicants compared to 248,848 last year. Factors contributing to this trend include a challenging NEET exam, a growing interest in technology, and high capitation fees for management quota seats in pure science courses. As of now, 239,871 have paid fees and 214,005 uploaded required documents. Experts like D. Valavan and Jayaprakash Gandhi highlight that many students prefer engineering due to the costs associated with management quotas in BSc programs. Interest in courses like computer science, AI, and data science is rising, influenced by generative AI advancements. However, concerns about future employability persist, requiring students to adapt and gain practical skills. Industry leaders emphasize the expanding job market in tech, as increased demand from global companies and investments in tech provide ample career opportunities for engineers.
Source link
Discover the Ultimate AI Tool: A Blend of Perplexity, NotebookLM, and ChatGPT – Here’s Why You Can’t Miss It!
Navigating the vast array of AI tools can be overwhelming, especially for students seeking effective research support. While major platforms like ChatGPT and Gemini are popular, they aren’t always the best fit for individual needs. One standout option is Logically (formerly Afforai), an AI workspace designed for both research and writing. It integrates features such as document creation, bullet points, and tables, similar to other platforms like Notion and Word. Logically enhances its usability with AI research tools, including a Semantic Scholar feature that searches a database of 200 million peer-reviewed papers, providing references and quotes. Its built-in chatbot can answer queries using advanced AI models, while allowing users to organize and manage their sources. Logically’s comprehensive functionality makes it particularly beneficial for students and researchers who need a centralized tool for managing projects and references effectively, streamlining their workflow significantly.
Source link