Google promotes AI as “enterprise-ready” with a stronger emphasis on verifiable information and less reliance on speculation.

Google promotes AI as “enterprise-ready” with a stronger emphasis on verifiable information and less reliance on speculation.



Google Cloud’s development platform, Vertex AI, is focusing on acquiring new skills to improve the accuracy of applications and services. Recently, Google announced that customers using Vertex AI will have the option to enhance the AI results of their services with specialized third-party data sets from vendors like Moody’s, MSCI, Thomson Reuters, and ZoomInfo. This feature is set to be available in the third quarter of this year.

In addition to this, Google is introducing several new features aimed at encouraging organizations to adopt Enterprise-Ready Generative AI Experiences. One such feature is “high fidelity mode,” which allows organizations to customize results generated from their own corporate data sets rather than relying solely on Gemini’s knowledge bank. This mode, compatible with Gemini 1.5 Flash, is now available for preview via the Vertex AI Experiment Tool.

Google is also enabling organizations to leverage their enterprise data sets to generate insights using Google AI models. Moreover, the platform is expanding its vector search capabilities to support hybrid search, allowing users to combine vector-based searches with text-based keyword searches for improved accuracy. This update is currently available in public preview.

Furthermore, Google is working on introducing a feature called “dynamic retrieval” to Vertex AI’s grounding with Google Search functionality. This feature will automatically choose whether to pull information from Gemini’s data sets or from Google Search, depending on the type of question and the need for frequently updated resources.

Overall, these updates to Vertex AI aim to enhance the accuracy and reliability of AI-generated information provided by applications and services. Google is striving to make its platform more effective for organizations looking to adopt AI technologies, ultimately reducing the frequency of models returning misleading or inaccurate results.

Article Source
https://www.theverge.com/2024/6/27/24187534/google-vertex-ai-developer-search-features-third-party-data