By Michael Scott
The ElevenLab’s new speech generator can now aid one in speaking another language without an accent.
The artificial intelligence arms race keeps accelerating, with new voice cloning limits occurring almost every day.
The most current development involves ElevenLabs, a San Francisco-founded startup that just revealed that their new model can imitate voices by talking effortlessly in 30 languages. This marks a significant growth from the previous eight that were earlier supported.
ElevenLabs Leverages Artificial Intelligence to Generate Speech
The firm utilized Lukeman Literacy, an autonomous publisher and literacy agency, as an illustration, clarifying that it develops several audiobooks yearly in several languages. Via an official blog post, ElevenLabs claimed that Lukeman’s team spent several weeks developing a single audiobook.
The post indicated that the process was time intensive given the need to identify the most suitable voiceover artist, arrange for a recording studio, and manage the post-production. Currently, the whole process takes less hours.
ElevenLabs Confirm Artificial Intelligence-Powered Doppelgangers with Multilingual Capabilities
ElevenLabs claims that the latest Multilingual v2 version delivers emotionally rich’ audio capable of capturing the natural speech’s nuanced inflections.
A user writes the text they intend to have spoken in the target language, and the artificial intelligence creates a coherent voiceover. The firm offers two major cloning alternatives, including a ‘VoiceLab’ for cloning explicit voices and a text-to-speech tool.
Users upload samples of speech to make a custom voice clone, which the artificial intelligence evaluates to create a copied form. The replicated voice can be altered to say all imaginable things. According to ElevenLabs, this current update indicates that artificial intelligence doppelgangers can speak smoothly in tongues like Malay, Swedish, and Arabic.
The improved linguistic abilities also overlap ElevenLab’s shift of its voice cloning technology out of beta testing. The firm seeks to market this tool for practical uses, for instance, narrating audiobooks.
ElevenLabs Faces Backlash for Misuse in Speech Generator
The likelihood of misuse impacts the business ambitions. In this case, deepfake audios result in users being susceptible to misinformation campaigns and fraud. Last year, ElevenLabs endured backlash following the exploitation of its platform to harass and imitate prominent persons.
The firm claims that despite stricter precautions being put in place since then, ethical concerns continue. A recent report indicates that a scammer could utilize artificial intelligence to mimic a loved one’s voice. All that was needed to attain realistic results was a few minutes of audio.
Meta Withholds Voicebox Citing Susceptibility to Misutilization
Major technology companies such as Meta encounter the same reproach for creating robust generative artificial intelligence without total transparency. Recently, the firm introduced Voicebox, an AI speech synthesis tool, and recognized its capability to enable deepfakes. Different from ElevenLabs, Meta avoided any release to the public owing to the risks of misutilization.
in spite of the concerns, quick progress in artificial intelligence voice cloning appears unrelenting. Mati Staniszewski, ElevenLab’s linguist, claimed that they hope to use AI to cover more voices and languages and eradicate the linguistic obstacle to content.
Ethical enactment is a major task since the line between worldwide misrepresentation and innovative communication methods is relatively thin. It is critical to tread carefully to prevent the universal village of voices from becoming a discordant Tower of Babel.