AI’s New Turing Test
A new Turing Test has been proposed to measure AI’s ability to exhibit human-like behavior. The test focuses on the machine’s ability to create a story with a coherent narrative, rather than just answering questions. This new approach aims to assess the AI’s understanding of complex human emotions, relationships, and social norms. The test involves the AI system reading a short story and then continuing it in a way that aligns with the characters, plot, and emotional tone. This new Turing Test is seen as a more nuanced way to evaluate AI’s understanding of human-like storytelling and could lead to advancements in creating more empathetic and engaging AI systems. Read more
White House Offers Prize Money for Hacker-Thwarting AI
The White House has announced a competition offering prize money to develop AI systems that can thwart hackers. The initiative is part of a broader effort to bolster the nation’s cybersecurity infrastructure. The competition aims to encourage the development of innovative AI solutions that can detect and prevent cyber threats in real time. The White House is collaborating with various agencies and private sector partners to ensure the success of this initiative. The goal is to create a more resilient cybersecurity ecosystem that can adapt to evolving threats and protect critical infrastructure. Read more
AI Bots to Seek Consent for Using Personal Material in News
A new AI system has been developed that seeks consent from individuals before using their personal material in news stories. This technology aims to address concerns about privacy and the unauthorized use of personal information in journalism. The AI system works by identifying personal material within a news story and then contacting the individual to seek their consent. If consent is not granted, the material is either not used or anonymized. This approach represents a significant step towards ensuring ethical journalism and protecting individual privacy in the digital age. Read more
Incorporating Human Error into Machine Learning
Researchers are developing a method to incorporate human uncertainty into machine learning systems. Many AI systems are programmed to assume that humans are always certain and correct, but real-world decision-making includes occasional mistakes and uncertainty. Researchers from the University of Cambridge, The Alan Turing Institute, Princeton, and Google DeepMind are working to bridge the gap between human behavior and machine learning. By training with uncertain labels, they found that these systems’ performance in handling uncertain feedback improved. This could help reduce risk and improve trust and reliability in AI applications, especially in safety-critical areas like medical diagnosis. Read more
Colleges Seek to ‘ChatGPT-proof’ Assignments
Educators are rethinking how they teach courses to “ChatGPT-proof” test questions and assignments, as AI chatbots like ChatGPT have become a source for cheating in college. Some instructors are returning to paper exams, while others are requiring students to show editing history and drafts to prove their thought process. The challenge of detecting AI-generated text has led to new questions for academics and a need to ensure that students understand how to do the work. The rise of AI chatbots has also impacted study habits, with some students using AI platforms for free instead of paid services. The shift back to paper-based tests and personalized writing assignments is seen as a way to ensure students learn the material and reduce cheating. Read more
Microsoft Unveils Next-Gen AI Solutions for Frontline Productivity
Microsoft has unveiled a suite of tools and integrations designed to empower frontline workers, harnessing the capabilities of generative AI to enhance efficiency and effectiveness. The innovative Copilot offering is central to this release, aimed at assisting frontline service managers and technicians by automating repetitive tasks. Microsoft estimates that over 60% of the 2.7 billion global frontline workers grapple with monotonous tasks. The new AI-driven tools, including Dynamics 365 Field Service integration and a new mobile experience, aim to alleviate the burden of essential yet exhaustive procedures, enabling swifter and well-informed choices. These advancements are seen as the first steps in infusing next-gen AI with productivity tools to address challenges such as repetitive tasks and burnout among frontline workers. Read more
