Sesame, a new startup founded by the same team behind the virtual reality company Oculus, has just released its conversational AI app for iOS. This move brings their vision of more natural, back-and-forth interactions with artificial intelligence to a wider public. Unlike the often clunky experiences with typical chatbots, Sesame aims to make these digital conversations feel more intuitive and human-like.
The founders, well-known for their work in virtual reality, are now tackling a different frontier: making AI agents better at understanding and responding in real-time. Think of it like moving from a simple walkie-talkie conversation, where one person speaks then the other, to a fluid phone call where interruptions and natural pauses are common. This approach could significantly change how we interact with digital assistants and customer service bots, making those exchanges far less frustrating.
This isn't just about making AI sound more polite. It's about designing the underlying technology, often referred to as an LLM (large language model, the tech behind ChatGPT), to anticipate and adapt to human conversational patterns. This involves processing speech and intent faster, allowing for interruptions and more nuanced responses. For everyday users, this means less waiting for a bot to finish its pre-programmed sentence and more direct, helpful interactions.
The launch of the Sesame app on iOS means anyone with an iPhone can now test this new style of AI conversation. While many companies are working on improving AI's conversational abilities, Sesame's pedigree from the Oculus team brings a unique blend of user experience and deep tech understanding. Their past success in making complex technology accessible suggests they could make a real impact on how we perceive and use AI in our daily lives.
What to watch next: How will users react to these more 'human' AI conversations? Will Sesame's approach set a new standard for how we expect AI to communicate, pushing other tech giants to adopt similar real-time, fluid interaction models?