ai voice

"Yuki Kaji is focusing on the AI voice project "Soyogi Fractal.""

"I completed this possibility experiment during the 7-day New Year's holiday in 2026."

"But the coolest part is that this means that you can use math to modify voices, average voices. You can swap gender, pitch, mix and match voices, and even create an emotion space!"

"I asked because I want to create some narrated versions of public domain stories as well as Conan the barbarian (it's still 2 years off from being in the public domain but come on that just splitting hairs)."

"Following the release of the AI podcasting feature in 2024, the internet praised how the podcasters sounded more human than expected."

"Former NPR host David Greene is suing Google after accusing the tech giant of stealing his voice for use in one of its AI-powered tools."

"The point is to limit the switching timing to the start of TTS playback."

"This tool enables users to generate professional-quality rap songs instantly just by entering ideas and themes in text."

"The ability to design voices with descriptive text is particularly exciting."

"AI voice technology company ElevenLabs' valuation reaches 1.7 trillion yen."

"AI voice startup ElevenLabs raised a $500M Series D led by Sequoia at an $11B valuation, taking its total funding to nearly $800M, and reports $330M in 2025 ARR"

"If you're spending time on text-to-speech conversion, utilizing AI voice generators is highly recommended."

"Basically: ElevenLabs-style voice cloning + NotebookLM-style podcast generation, but local."

"You can now finnally create your own AI voice. It lets you just type a description like "calm female voice with British accent" and it generates a voice for you."

"10 years ago, remember that wave of hardware?"

"Qwen3-TTS is generating buzz, with people asking, 'Do we even need ElevenLabs?'"

"This article explores the capabilities of Qwen3-TTS, offering a complete guide to understanding and using this cutting-edge voice cloning AI."

"Hume AI’s CEO, Alan Cowen, will join Google DeepMind along with several top engineers as part of a major licensing deal."

"Anker and Feishu's cooperation can be seen as a test of its AI recording hardware market. It can use Feishu's mature software capabilities to make products, and it can also verify its own R&D and mass production capabilities in this type of hardware product."

""Bringing together our engineering teams and our respective technologies will drive faster value for customers and more intuitive ways of working with AI," said Amit Zavery, president, chief operating officer, and chief product officer at ServiceNow."

"HHO is launching the AI voice recorder "TALIX & DingTalk A1" on January 17th, a product jointly developed by digital product brand "TALIX" and intelligent work platform "DingTalk.""

"I have designed it for massively improved stability and audio quality over the original model. ... I have trained Soprano further to reduce these audio artifacts."

"The actual problem is that when you don't give ChatGPT enough constraints, it gravitates toward the statistical center of its training data."

"turns out AI voices are weirdly perfect. like 0.002% timing variation vs humans at 0.5-1.5%"