Youssef “Benathman” Idelcaid’s Post

One of our favorite people to talk to about how humans think and relate to AI—Steven Sloman, professor of cognitive, linguistic, and psychological sciences at Brown University. Add this to your weekend listening and hear from one of the original discoverers of our intuitive versus deliberative cognition aka system 1/system 2 and what this may mean for large language models and AI in general. #generativeai #cognitivescience #artificialintelligence #artificialgeneralintelligence https://lnkd.in/gZxNVPhX

Spoken dialogue plays a crucial role in human-AI interactions, necessitating dialogue-oriented spoken language models (SLMs). To develop versatile SLMs, large-scale and diverse speech datasets are essential. Additionally, to ensure hiqh-quality speech generation, the data must be spontaneous like in-wild data and must be acoustically clean with noise removed. Despite the critical need, no open-source corpus meeting all these criteria has been available. This study addresses this gap by constructing and releasing a large-scale spoken dialogue corpus, named Japanese Corpus for Human-AI Talks (J-CHAT), which is publicly accessible. Furthermore, this paper presents a language-independent method for corpus construction and describes experiments on dialogue generation using SLMs trained on J-CHAT. Experimental results indicate that the collected data from multiple domains by our method improve the naturalness and meaningfulness of dialogue generation. #AI #SpeechModels #DialogueGeneration #SpokenLanguage #LargeScaleDatasets

Voice AI's Role in Overcoming Language Barriers in Cold Calling Language barriers can often pose significant challenges in various aspects of life, and the world of cold calling is certainly no exception. However, thanks to the advancements in technology, a solution has emerged in the form of Voice AI. This powerful tool has revolutionized communication by [breaking down linguistic barriers](https://lnkd.in/e-7zm5gi), enabling seamless conversations across languages. In this article, we will explore the capabilities of Voice AI and how it plays a crucial role in overcoming language barriers in cold calling. https://lnkd.in/eeY9VDmd

Conversation powered by a new form of empathic AI Hume's EVI is an API powered by our empathic large language model (eLLM), which understands and emulates tones of voice, word emphasis and more to optimize human-AI interaction. A universal voice interface: a single API for transcription, frontier LLMs, and text-to-speech. End-of-turn detection: uses your tone of voice for state-of-the-art end-of-turn detection, eliminating awkward overlaps. Interruptibility: stops speaking when interrupted and starts listening, just like a human. Responds to expression: understands the natural ups and downs in pitch & tone used to convey meaning beyond words. Expressive TTS: generates the right tone of voice to respond with natural, expressive speech. Aligned with the application: learns from users' reactions to self-improve by optimizing for happiness and satisfaction. #inceptionx #evi #humeai #voiceinterface #emotionalintelligence #artificialintelligence

Spoken dialogue plays a crucial role in human-AI interactions, necessitating dialogue-oriented spoken language models (SLMs). To develop versatile SLMs, large-scale and diverse speech datasets are essential. Additionally, to ensure hiqh-quality speech generation, the data must be spontaneous like in-wild data and must be acoustically clean with noise removed. Despite the critical need, no open-source corpus meeting all these criteria has been available. This study addresses this gap by constructing and releasing a large-scale spoken dialogue corpus, named Japanese Corpus for Human-AI Talks (J-CHAT), which is publicly accessible. Furthermore, this paper presents a language-independent method for corpus construction and describes experiments on dialogue generation using SLMs trained on J-CHAT. Experimental results indicate that the collected data from multiple domains by our method improve the naturalness and meaningfulness of dialogue generation. #AI #SpeechModels #DialogueGeneration #SpokenLanguage #LargeScaleDatasets

Curious about the future of language professionals in the age of AI? Mark your calendar for Nov 22, 5 PM (CET), and tune in to learn from Arle Lommel as he delves into the transformative power of Generative AI. Register now! https://lnkd.in/emh4fKVy #LanguageProfessional #AI #FutureofAI

👏Another excellent video on language AI definitely worth watching. If still necessary it demonstrates that language technology is everywhere because language is everywhere in our life. ➡A few points I really like in this discussion ⚡Universal language inclusivity and accessibility will make AI truly inclusive ⚡Multimodal AI will transcend language barriers beyond what we are used to ⚡As language technology becomes more capable responsible AI becomes more important ⚡Language AI will enable us to interact not only with the digital/online world but also with the real world 🚀More challenges and opportunities ahead... Also, I like this simple definition of language that is enough to justify the central role of language technology and intelligence⬇ "Every language is a universe within itself, a complex ecosystem of sounds, grammar, and meaning that has evolved over millennia. It's a reflection of a culture, a history, and a unique worldview."

Recent work confirming perspectives in my 2010 book The Learning Layer and subsequent blog posts such as this one: https://lnkd.in/gbXUXY76

Is this the summer break where you planned to catch up on what Large Language Models are and how they apply in life sciences? Here are a few pages to help de-mystify the topic in 15 minutes and free from any form fills. #AI #LLMs #Generative AI #lifesciences #TRUST https://lnkd.in/d827TTc6

#MIT researchers make the case that language is a tool for communication, not for thought … What is the implication to Generative AI and Artifical General Intelligence under this light? The knowledge based GenAI will remain as an effective knowledge dissemination tool with simple reasoning ability learned in books… the true creativity from human ingenuity and brilliance arising from human thoughts is yet to be deciphered … Having said that, knowledge dissemination and simple reasoning can unlock significant value… when integrated with domain specific data, tech, skills and work processes …they can help accelerate human ingenuity … https://lnkd.in/gsiY6YK7