Amazon Project Olympus is directing substantial investments into the development of an expansive language model (LLM) named Amazon Project Olympus Pioneering AI Advancements with a Massive Language Model positioning itself to compete with top models from OpenAI and Alphabet, insiders familiar with the matter have disclosed to Reuters.
The Potential of Amazon Project Olympus
The rumored 2 trillion parameters of “Olympus” could potentially establish it as one of the largest models globally, surpassing the one-trillion-parameter GPT-4 by OpenAI. This clandestine project was disclosed by sources opting for anonymity due to its confidential nature. Amazon, queried about the matter, chose to remain silent, with The Information first reporting on the project earlier this week.
Leading this innovative initiative is Rohit Prasad, former head of Alexa, who now reports directly to CEO Andy Jassy. In his role as Head Scientist of Artificial General Intelligence (AGI) at Amazon, Prasad has assembled a specialized team, bringing together researchers from the Alexa AI and Amazon science teams to concentrate on model training, unifying AI endeavors across the company with dedicated resources.
Previous Forays and Collaborations
Amazon’s previous forays into model training include smaller projects like Titan, alongside partnerships with AI model startups such as Anthropic and AI21 Labs. These collaborations extend to Amazon Web Services (AWS) users, reflecting Amazon’s commitment to pushing the boundaries of AI innovation.
The e-commerce giant envisions that cultivating in-house models will elevate the allure of its AWS offerings, particularly for enterprise clients seeking access to top-tier models. Despite the covert nature of the project, Amazon has not disclosed a specific timeline for the release of the “Olympus” aka Amazon Project Olympus model.
Acknowledging the Importance of Large Models
Language models with extensive parameters, like LLMs, are the cornerstone of AI tools capable of learning from vast datasets, generating responses imbued with human-like nuances. The financial commitment to training larger AI models is acknowledged by Amazon executives, who, during an April earnings call, emphasized increased investments in LLMs and generative AI. This strategic pivot aligns with a simultaneous reduction in spending on the fulfillment and transportation aspects of Amazon’s retail business.
Read More (AI)
