TL;DR:
- Oracle Corp. collaborates with Cohere, a Toronto-based startup specializing in large language models (LLMs), to develop a new cloud service.
- The partnership aims to make it easy for enterprise customers to train their own customized LLMs while ensuring data privacy and security.
- Oracle’s internal teams have already utilized the Cohere AI cloud service for specialized LLMs in healthcare.
- Larry Ellison, Oracle’s founder and CTO, emphasizes the company’s expertise in handling large amounts of data securely, giving it an edge over competitors.
- Oracle Cloud is a multi-billion-dollar business for AI workloads, with various companies utilizing it for LLM training.
- While Oracle is an established player in AI, it faces strong competition from hyperscalers like AWS, Google, and Microsoft in the generative AI services market.
Main AI News:
In a recent earnings call, Larry Ellison, the founder and chief technology officer of Oracle Corp., unveiled an exciting development in the company’s portfolio: a groundbreaking cloud service powered by generative AI. This new venture is being undertaken in partnership with Cohere, a leading Toronto-based startup renowned for its expertise in constructing and training large language models (LLMs).
Ellison, while highlighting Oracle’s remarkable growth in the cloud business sector, confirmed the collaboration with Cohere, emphasizing their joint efforts to facilitate enterprise customers in training their own customized LLMs using proprietary data, all while prioritizing data privacy and security. The news of this partnership comes as no surprise, given the close relationship between the two companies. Oracle recently participated in Cohere’s impressive $270 million series C funding round, which valued the startup at a remarkable $2.2 billion. Notable investors in this round included the likes of Nvidia Corp., Salesforce Ventures, Deutsche Telekom AG, and SentinelOne Inc.
During the earnings call, Ellison stated, “Cohere and Oracle are working together to make it very, very easy for enterprise customers to train their own specialized large language models while protecting the privacy of their training data. Over the next few years, lots of companies are going to train their own specialized large language models.“
Moreover, Ellison revealed that Oracle’s internal application development teams are already leveraging the Cohere AI cloud service, powered by Oracle Cloud Infrastructure (OCI). The company utilized its private data to fine-tune and expand the existing Cohere LLMs, leading to the creation of two specialized models: one tailored to medical professionals and another designed for first responders. This move aligns with Oracle’s strategic presence in the healthcare sector, following its acquisition of healthcare giant Cerner in 2022.
Ellison highlighted the significance of specialized large language models in enhancing the efficiency of highly trained professionals, enabling them to optimize their valuable time. While renowned for its database technology, Oracle’s prominence in the healthcare industry positions it as a key player in driving advancements in this field.
Despite this exciting collaboration, Oracle is not a newcomer to the world of AI. Ellison emphasized during the earnings call that Cohere utilizes Oracle Cloud for training LLMs. Oracle’s extensive experience and expertise in securely and efficiently handling vast amounts of data give the company a competitive edge over its rivals. Adept AI Labs, a company that raised $350 million for its generative AI service, and Nvidia, through their cloud AI partnership, are among the entities that publicly acknowledge their use of Oracle Cloud for LLM training. Ellison proudly declared that Oracle Cloud has already established itself as a multi-billion-dollar business for AI workloads.
However, while Oracle’s progress in the cloud domain is notable, it still trails behind the industry’s major players, often referred to as the “big three hyperscalers.” Amazon Web Services (AWS), for instance, introduced its Bedrock generative AI services in April, while Google showcased an array of updated services and models at its recent I/O conference. Microsoft benefits from its close alliance with OpenAI. These companies have their own generative AI services, putting them at the forefront of the industry.
Conclusion:
Oracle’s partnership with Cohere and their development of a state-of-the-art AI cloud service demonstrate the company’s commitment to advancing in the AI landscape. By leveraging its expertise in data management and security, Oracle aims to empower enterprise customers to train specialized large language models effectively. However, they face fierce competition from major hyperscalers who have already established their own generative AI services. Nonetheless, Oracle’s strategic moves position them as a significant contender in the evolving market, with the potential to shape the future of AI-driven services.
