All articles tagged with #voice recognition
Anthropic has introduced a new optional setting for Claude users to share voice recordings for AI model training. The feature, which appears as a prompt during voice interactions, is disabled by default and separate from existing text-based training permissions. Users can enable or disable the option in privacy settings and delete data at any time.
Smart TVs increasingly use onboard microphones and automatic content recognition (ACR) to identify what you’re watching and may be paired with wake-word processing that stays local, with some features labeled opt-in. While manufacturers argue these tools only activate when prompted, researchers warn they could be hacked or misused to spy, and most brands also collect data for ads and analytics. Reducing tracking means diving into privacy and ad settings to disable microphone use and ACR-related features, but a truly private setup is hard to achieve in the modern connected world—as with many devices, complete anonymity isn’t realistic, so mitigate what you can or opt for a simpler setup.
LG says its smart TVs only record audio when the user actively uses a voice command (via the remote button or the ‘Hi LG’ wake word) and that such data is deleted if the wake word isn’t recognized; the company emphasizes explicit opt-in controls and separate handling of advertising/data-collection features, even as a recent investigation questioned potential background listening and local buffering.
Samsung’s smart TVs collect viewing data, voice commands, and use this information for advertising and analytics by default (ACR and related features). To disable, go to Settings > Privacy Choices and turn off Viewing Information Services (ACR), disable Interest-Based Advertising, and switch off Voice Recognition Services. Doing so reduces data sharing with Samsung and targeted ads, though it won’t affect core TV performance.
Research shows that specific parts of the human brain respond distinctly to chimpanzee vocalizations, indicating deep evolutionary roots of voice recognition that predate human language and suggesting shared neural mechanisms with our closest primate relatives.
Handy is a free, easy-to-use speech-to-text application that leverages advanced AI models like Nvidia's Parakeet and OpenAI's Whisper to convert spoken words into text accurately, offering a simple setup and customizable options for users on Windows, macOS, and Linux.
After three months of using Alexa+, the upgraded AI assistant from Amazon, the author highlights its more natural, conversational interactions, improved smart-home controls, a redesigned Echo Show interface, better management of lists and reminders, and new visual recognition capabilities, though it still falls short in fully autonomous actions. Overall, Alexa+ shows significant progress but remains a work in progress for complete automation.
Waze has expanded its 'Conversational Reporting' feature, allowing users to report road hazards using natural voice commands, making it easier to share real-time updates while driving. The feature, initially in limited beta, is now being broadly rolled out, though some bugs and notification issues remain.
A Malvern-based startup, FarX, has developed an AI-powered security technology that combines voice and facial recognition to enhance online security and detect fraud, with plans to incorporate emotional recognition to improve customer service interactions. The company has secured patents in the UK and US and is already used in the banking industry.
Taco Bell is reconsidering its use of AI at drive-throughs after mixed results and viral moments, with plans to allow franchisees to decide when to deploy voice AI or rely on human staff, emphasizing a tailored approach to technology adoption.
Microsoft's Windows future will be more ambient, pervasive, and multi-modal, heavily integrating AI to enable context-aware, voice-driven interactions, transforming the OS into an agentic assistant that seamlessly combines local and cloud capabilities, with significant changes expected in user interfaces and experience over the next few years.
The article reviews Alexa+'s preview release, highlighting its improved conversational abilities and integration with third-party services, but also its limitations like inconsistency, incomplete task execution, and AI's imperfect understanding, raising philosophical questions about AI's human-like qualities.
The reviewer tests Amazon's new Alexa Plus, which no longer requires specific voice commands and features an ambitious but still limited agentic AI that aims to act on users' behalf. While it understands natural speech better and can manage tasks like booking tickets or fixing appliances, its capabilities are currently restricted by reliance on specific services and lack of web-wide browsing, making it not yet a full replacement for a personal assistant.
SoundHound AI (SOUN) stock continues to surge as its voice recognition and response technology gains traction, with an upcoming presentation at Nvidia's conference contributing to the excitement. The company's focus on AI applications in marketing and transcription, along with its investment from Nvidia, has driven the stock's recent gains, despite concerns about competition and voice cloning. SoundHound's potential in connecting database computing to voice recognition presents opportunities for further growth in the AI revolution.
Apple, Microsoft, Amazon, Alphabet, and Nvidia have all made significant investments in voice-recognition software, reflecting the growing interest in AI-powered voice assistants. Nvidia's recent investment in SoundHound AI has caused the stock to surge, highlighting the potential for growth in this sector. With the market for voice-recognition technology projected to reach $50 billion by the end of the decade, SoundHound AI, despite its current small size, could emerge as a major player in the long run.