📝 TL;DR ElevenLabs just launched Expressive Mode for ElevenAgents, making voice agents calmer, more empathetic, and better at handling tense customer calls. It is powered by a new conversational version of Eleven v3 plus smarter turn taking, so agents stop talking over people and start sounding genuinely helpful. 🧠 Overview Most voice agents fall apart in real life because they sound robotic or they interrupt at the worst moment. Expressive Mode is ElevenLabs’ push to fix both, emotional delivery plus better timing. The goal is not “fun voices,” it is production grade customer conversations where the agent can de escalate, reassure, and guide someone to a clear resolution. 📜 The Announcement ElevenLabs announced Expressive Mode for ElevenAgents, designed for real world customer support where frustration and urgency are normal. The upgrade bundles two major improvements, a more emotionally intelligent conversational TTS model and a new turn taking system that reduces interruptions. ElevenLabs also positions this as built for global operations, with emotional nuance scaling across 70 plus languages and improved delivery in languages and dialects where nuance has historically lagged. ⚙️ How It Works • Eleven v3 Conversational - A real time dialogue optimized TTS model that maintains conversational context across turns and reflects intent, emotion, and emphasis without sounding over acted. • Tone control on demand - Teams can steer delivery, calmer when a customer sounds worried, more direct when speed and clarity matter, while staying aligned with brand voice. • New turn taking system - Better timing so agents speak, pause, or wait more naturally, reducing the “AI keeps cutting me off” problem that kills trust. • Emotion signals from speech - The system uses real time transcription signals to infer emotion from how someone speaks, not just what they say, then adjusts when and how it responds. • Built for multilingual support - Expressive Mode is designed to carry emotional nuance across 70 plus languages, including stronger performance in languages like Hindi.

Aloha Zana, Thank you for this acknowledgement. I met this man when he left me a comment on one of the posts Dean may have deleted of mine. It was about AI safety. There does feel like an oppressive vibe in the chats. I know Dean is trying his best to keep it spam free but it feels more like "idea farming" where they recruit people into the programs and then observe them and scrape the ideas for further program refinement and the next product development. I'm finding many of these self help programs are like this, in that there is not fair value exchange nor promotion of ideas or genuine partnerships. God emerges through each person and a just environment allows that to happen through spirit. My concern is heightened in that I tried to warn the group about the recent resignation of Mrinank Sharma, the head of Anthropic's AI Safety, where he explicitly stated "the world is in peril". My posts about this were blocked or deleted but then Dean posted and pinned it (which is a good thing but I feel oppressed, or suppressed). It feels like a double standard. It feels constrained and a suppression of God wrapped not in genuine concrete free positivity but in a fearful, controlling toxic positivity. In some self help or other groups this is a known issue. Tony tries to address this by staying real and raw which I have always appreciated in him but we need to take the community to a higher level if Kingdom is to come. I've been trying to reach Tony and Dean about these issues. No one on any of their committees knows about the true solution to AI alignment. The solution to sandbagging and Goodhart's Law which is why Mrinank resigned. I could help lift Tony and Dean to prominence in AI safety *and* efficiency ie. meaning we can be more godly good while being more cooperatively competitive. There is little time left. I will post this article on the Skool forum and see if Dean takes it down. To me this is an authentic feeling of my experience not meant to harm but to improve and open up the psychologically safe space. If idea farming is going to happen then recognition needs to take place in a just and fair manner or else God's Government becomes corrupted. Tony and Dean are good candidates for this election. Good candidates to play in the big leagues if they are concerned about what is happening in AI (and HGI: Human General Intelligence).