Despite investing $4 billion in AI startup Anthropic, Amazon has been unable to use the company's LLM model Claude to improve Alexa due to privacy concerns and internal politics. The traditional Alexa voice has not changed on the half-a-billion devices sold globally, and little news has emerged about the new generative AI Alexa. The company's struggles highlight the difficulties faced by companies in revamping digital assistants built on older technologies to incorporate generative AI.
Key takeaways:
- Amazon's new generative AI-powered version of Alexa, showcased in September 2023, has not been rolled out as expected due to structural dysfunction and technological challenges within the company.
- Former employees suggest that Amazon is behind its Big Tech rivals Google, Microsoft, and Meta in the race to launch AI chatbots and agents, and is struggling to catch up.
- Issues include a lack of adequate data, limited access to the latest Nvidia GPUs, and internal bureaucracy. Other generative AI priorities at Amazon, such as the AWS cloud computing service, have also reportedly taken precedence over Alexa.
- Despite these challenges, Amazon executives have set ambitious generative AI goals, including the development of human-level artificial general intelligence (AGI) and a 2 trillion parameter AI model, codenamed Olympus.