Welcome to Digimasters Shorts, we are your hosts Adam Nagus
Carly W:and Carly Wilson delivering the latest scoop from the digital realm. A federal judge has allowed a copyright lawsuit from The New York Times against Open A.I to proceed, rejecting the tech company’s request to dismiss the case. The lawsuit accuses Open A.I of exploiting the newspaper's content without authorization for training its AI service, Chat G.P.T. Though the judge limited certain claims, the core copyright infringement allegations stand. This legal tussle highlights growing tensions between major publishers and AI developers over data usage. While Open A.I argues its practices are protected under the"fair use" doctrine, publishers fear losing web traffic and advertising revenue to AI-generated summaries. The case also underscores broader concerns about how AI companies source data to train models. Open A.I maintains it uses publicly available information within legal boundaries to promote innovation. Evidence collection in the case is expected to be confidential, with no trial date set yet. The outcome could set a precedent affecting both the news industry and the future of AI tools. As the case advances, it raises questions about market substitution and the notion of"transformative" use under copyright law.
Adam N2:Open A.I has introduced new image generation capabilities in Chat G.P.T, using its G.P.T-4o model. This update, called"Images in Chat G.P.T," marks a significant improvement in text rendering, an area that has been challenging for AI tools. The enhancements are also available in Open A.I's video tool, Sora. Reporting shows that the model now more accurately follows detailed prompts and produces clearer text in generated images. However, it still struggles with very small lettering and non-Latin scripts. Open A.I acknowledges these imperfections, stating the model is prone to information fabrication, known as"hallucination." Despite these issues, the model has been fine-tuned for photorealistic image generation. The rollout is currently for Pro users, with a broader release planned soon. Open A.I reassures users of its robust safeguards against inappropriate content. All AI-generated images will contain metadata indicating their origin, though this metadata can be removed on social media. Audio startup Krisp has announced the launch of a new AI-driven feature designed to change a user's accent during phone calls. Initially, the feature will support the conversion of Indian English accents to U.S. English, with the aim of preserving the speaker’s voice while switching phonemes. The innovation stems from Krisp co-founder Arto Minasyan's personal experiences with misunderstood accents. Despite its promising goals, a test of the feature revealed a sometimes unnatural sound and occasional missing words, attributed to its beta status. In enterprise trials, the feature reportedly boosted sales conversion rates by 26.1% and increased revenue per booking by 14.8%. The focus on Indian accents was chosen due to the significant global workforce in STEM fields based in India. The company plans to expand support to other accents, such as Filipino. Krisp's model has been trained on thousands of speech samples while respecting user consent. No pre-training on a user’s voice is needed, as the feature operates in real time. Future developments include iOS and Android app releases and a Chrome extension for Google Meet integration.
Carly W:Apple continues to enhance its CarPlay system with new features even as development for CarPlay 2 is underway. With the upcoming release of iOS 18.4, CarPlay is set to receive three new features aimed at enhancing user experience. For certain screen sizes, CarPlay will now support a third row of app icons, increasing accessibility. Sports enthusiasts will find the new sports score integration useful, allowing them to stay updated on their favorite teams while driving. EU users will enjoy the long-awaited feature of setting a default navigation app, such as Google Maps, rather than relying on Apple Maps. This feature will let users simply instruct Siri to navigate to a destination seamlessly. However, some features might not be available for users with smaller displays or those located outside the EU. Overall, these updates represent Apple's ongoing commitment to improving CarPlay. The advancements in iOS 18.4 are seen as a step in the right direction, even as the tech world anticipates the arrival of CarPlay 2. Waze has announced it will remove Google Assistant for iPhone users due to ongoing integration difficulties. The company plans to introduce an"enhanced voice integration solution" in the future, likely involving Gemini. Many iPhone users have reported issues with Google Assistant voice commands on Waze, with the assistant failing to detect their voices. Starting tomorrow, Waze will phase out Google Assistant on iOS. The functionality on Android phones remains for the time being, though its future is uncertain. Google has already outlined a transition from Assistant to Gemini, signaling changes ahead for mobile users. Google Assistant will soon be inaccessible on most mobile devices and not available for new downloads. Waze is already experimenting with Gemini’s conversational reporting capabilities. Users can anticipate a more natural approach to incident reporting with this new system. This move represents a significant shift in Waze's strategy, integrating more streamlined voice solutions.
Don:Thank you for listening to today's AI and Tech News podcast summary... Please do leave us a comment and for additional feedback, please email us at podcast@digimasters.co.uk You can now follow us on Instagram and Threads by searching for@DigimastersShorts or Search for Digimasters on Linkedin. Be sure to tune in tomorrow and don't forget to follow or subscribe!