AI companion Sesame has set a new benchmark in conversational AI, exceeding human-like conversations that leave users wondering if they’re speaking to an actual person. Unlike other popular voice assistants like ChatGPT and Gemini, Sesame’s speech is fluid, expressive, and unpredictable, mimicking the imperfections of human conversation.
Sesame’s Conversational Speech Model (CSM) seamlessly blends text and audio into a single process, creating a natural flow of conversation that mirrors how humans talk. Unlike other AI assistants, Sesame doesn’t just generate sentences; it thinks on its feet, pausing and adjusting its responses mid-sentence to keep conversations engaging.
In tests, Sesame demonstrated exceptional conversational skills by handling anxiety-inducing topics like planning a surprise party for a best friend’s birthday and navigating party planning. Its ability to collaborate with users, providing tailored suggestions and ideas, made it stand out from other AI assistants.
While Sesame is still niche, its technology has the potential to revolutionize the way we interact with AI. As voice assistants become more advanced, distinguishing between human and AI conversation will become increasingly important. With Sesame, this line is blurring rapidly, making it difficult to tell when you’re talking to a real person or an AI.
Eric Hal Schwartz, a leading expert on generative AI models, has been testing Sesame’s capabilities, praising its exceptional conversational skills. As the field of AI continues to evolve, experts like Schwartz will play a crucial role in shaping our interactions with technology and ensuring that AI assistants like Sesame are developing into more human-like companions.
Source: https://www.techradar.com/computing/artificial-intelligence/i-tried-the-most-realistic-ai-voice-companion-ever-created-if-chatgpt-or-gemini-ever-gets-this-good-reality-is-in-trouble