Tech News
← Home  ·  All topics

Voice

159 GoKawiil briefs on this topic · RSS for voice

Spotify adds new AI DJ voices and a low-chatter mode for Premium users

Spotify announced an upgrade to its AI DJ feature, letting Premium subscribers choose among several new voices with distinct personalities in English and Spanish, alongside a quieter mode with less commentary between songs. The update includes seven English voices with varied accents and two new Spanish voices, while the original 'X' voice remains available.

Otter.ai adds 27 languages, debuts mid-call language-switching transcription

Otter.ai announced it is expanding language support by 27 new languages, bringing its total to 31 once rollout finishes. A new multilingual transcription feature will automatically detect when speakers switch languages mid-conversation and keep transcribing without manual setting changes. Thirteen languages, including Mandarin, Korean and German, are live now, with 14 more such as Arabic, Hindi and Swedish coming later.

Google explains how to enable personal results on Gemini-powered Google Home speakers

Google is rolling out a Gemini-driven update to Google Home speakers, adding more conversational understanding and improved Voice Match and personal results features. Users can enable personal results by setting up Voice Match in the Google Home app, then turning on the feature under Google Assistant settings for chosen devices.

Raycast launches Dictate to AI, pairing voice commands with screen context

Raycast released a new command called Dictate to AI that lets users speak requests to its AI assistant while automatically pulling in context from whatever app is active on screen, building on its existing Screen Awareness feature. The update, part of Raycast v2.11, also adds a Dictation AI extension, a Mac menu bar AI usage tracker, a Contacts AI extension, expanded free-tier Notes support, iOS keyboard writing tools, and new Screen Awareness commands for Windows.

Meta rolls out voice chat for Muse AI assistant to all US users

Meta has launched a voice feature for its Muse AI agent, letting US users make phone calls or video chats with the assistant. The update offers hundreds of voice options and is now available to the general public rather than a limited test group.

Sonos opens early access to Sonos 27voice assistant on November 6

Sonos announced that its new voice assistant, called Sonos 27voice, will open for early access testing on November 6 as part of this week's software update. Users must request access through the Sonos app and be approved to try it ahead of a planned full release in 2027.

Startup Subvocal opens pre-orders for radar-based silent-speech glasses

Subvocal has begun taking pre-orders for a wearable that uses radio signals, rather than cameras or electrodes, to detect mouth and tongue muscle movement and convert silently mouthed words into text. The company says shipments are expected to begin in Q1 2027. The approach draws on prior research from MIT Media Lab's AlterEgo project, which used electrodes instead of radar.

Sonos adds System Health diagnostics to iOS app, more features coming to both platforms

Sonos has rolled out a new System Health page inside its iOS app that gives users a consolidated view of their speaker systems, including network status, room counts, and detection of 'Double Network Address' issues that cause dropouts and missing-speaker errors. The company's blog post also previewed upcoming updates for both iOS and Android, including volume presets, expanded miniplayer controls, and spoken voice prompts for the Ace Ultra headphones.

Sonos app update adds Presets, opens US waitlist for AI voice assistant

Sonos has released an app update introducing Presets, which let users save a specific volume and room combination for one-tap playback from the homescreen. The update also opens a US waitlist starting November 6th for Sonos' new 27voice AI assistant, with access beginning November 12th, letting users control playback, group speakers, set timers, and make natural-language requests.

Why recorded voices sound different to their owners, explained

A speaking coach and former TV producer explains a common phenomenon: people often recoil when they hear playback of their own voice. The reason, according to the coach, is that we normally hear our voice partly through bone conduction inside the skull, which adds lower frequencies that a recording cannot capture, making the recorded version sound unfamiliar and often less pleasant.

Openreach confirms January 31 shutdown of UK's analog phone network

Openreach, BT Group's network arm, told telecoms providers this week that the legacy PSTN copper phone network will be switched off on January 31, 2027, with no further delays. Customers still relying on old landline connections—including some point-of-sale systems, alarms and telecare devices—will need to move to Digital Voice services run over broadband routers.

A/SIDE details pricing, encryption scope and new voice-caption feature

A/SIDE's FAQ outlines how the private photo-sharing and messaging app works: free accounts get 30 days of feed and message history, while Pro Individual ($20/year) and Pro Family ($60/year) plans extend that history. Version 1.6 adds voice captions up to one minute on photos, and the company states that new direct and group chats are end-to-end encrypted, though feed posts and some legacy conversations are not.