GoKawiil
Tech News
Wispr Advanced Interfaces Lab has released Canto, a new speech recognition model built specifically for real-time dictation in noisy, everyday conditions rather than clean studio audio. In tests using 10 hours of real Wispr Flow dictations from over 2,300 speakers, Canto posted the lowest word error rate compared with models from Google, OpenAI, AssemblyAI, and Deepgram.
Apple devices include a toggle called Improve Siri & Dictation that, when enabled, allows the company to save and analyze transcripts and audio from Siri interactions, dictation, and translation requests. Many users may have turned this on during initial device setup without fully realizing what it permits, and it can remain active indefinitely without their awareness.
Jexxa is a new dictation tool that runs its speech recognition model entirely on the user's device, converting speech to text in any application while holding a key to talk. It avoids uploading audio or transcripts, sending only subscription and update checks, and can correct and learn from misrecognized names or jargon over time.
With today's iOS 27 rollout of Siri AI, Apple has replaced its old 'Improve Siri and Dictation' opt-in with a new setting called 'Improve Siri and Apple Intelligence.' Choosing 'Share Audio and Text' now lets Apple use voice and text interactions from Siri, Dictation and Translate to train its Foundation Models, not just refine Siri as before. Even users who previously consented will be prompted again during setup to confirm their choice under this expanded scope.
Startup Voicebox, led by CEO Karan Gupta, lets customers give feedback by speaking into their phones after scanning a QR code or tapping an NFC chip, with voice automatically transcribed and analyzed for sentiment on a company dashboard. The company has partnered with airport terminals and was used at a Lena Dunham book event, where attendees left spoken reactions instead of typed reviews.
ZDNET tested the Comulytic Note Pro, a compact AI-powered voice recorder shaped like a stack of credit cards that pairs with an app to transcribe and organize meetings, memos, and dictation. The reviewer praised its microphone quality, battery life, and unobtrusive design, giving it 4 out of 5 stars overall.
FnScribe is a new open-source macOS menu-bar app that transcribes speech locally on the device and inserts text directly into whatever app is focused. Users trigger it with a push-to-talk shortcut (the fn/Globe key) or a hands-free mode, and it includes smart cleanup features like filler-word removal, punctuation commands, list formatting, and voice-based correction phrases like 'scratch that.'