Search results for smart speaker
Telecommunications & Media
How Amazon built Alexa and Echo
Amazon took years of persistent experiments and trial and error to build a new voice-based computing experience. The Verge examined the internal process behind the creation of the Echo smart speaker and Alexa voice assistant. Jeff Bezos had long believed computing would move from keyboards and screens to voice. Engineers faced challenges in understanding speech and context, and in improving speed and accuracy. Competition and an unusual quiet launch strategy also shaped development and early reception.
AI & Enterprise
OpenAI develops voice model that adjusts responses in real time during conversation
OpenAI is developing an interactive voice model that can adjust responses in real time during a conversation, The Information reported. ChatGPT’s advanced voice mode works on a turn-based system that waits for a user to finish speaking, and it cannot change a response once generation begins. The new model, called BiDi, processes speech continuously and can adapt when interrupted. The prototype is not yet complete and may slip beyond the second quarter, the report said.
AI & Enterprise
Engineering AI gains momentum, expanding into chip design and 3D modelling
Industry efforts to move beyond large language models into engineering and physical AI are gaining traction, spanning semiconductor design, 3D modelling of products and robot operations. Cooperation among companies is increasing as interest in AI-led manufacturing innovation grows, and startups are accelerating their moves. The article rounds up recent product launches, investments and partnerships, including tools for design, automation, coding, enterprise productivity and AI security in South Korea and abroad.
