Google Cloud introduces new Vertex AI capabilities to enhance the accuracy and reliability of enterprise AI applications, addressing concerns about misleading or inaccurate outputs from generative AI models.
Key features for enterprise-ready AI: Google Cloud is rolling out new features for its Vertex AI platform to help organizations build more accurate and trustworthy AI services:
- Grounding with third-party datasets: Customers can now leverage specialized datasets from providers like Moody’s, MSCI, Thomson Reuters, and ZoomInfo to improve the accuracy of their AI models’ outputs, with this feature becoming available in Q3 2024.
- High-fidelity mode: Organizations can use their own corporate datasets as the primary information source for generated outputs, powered by a specialized version of Google’s Gemini 1.5 Flash model, currently available in preview through Vertex AI’s Experiments tool.
- Hybrid search for Vector Search: The Vector Search feature, which enables image searches based on similar graphics, now supports combining vector-based searches with text-based keyword searches for improved accuracy, available in public preview.
Addressing concerns about AI accuracy: These updates aim to tackle the issue of AI models providing misleading or inaccurate information, which has been a concern with Google’s AI-based search features:
- Google’s AI Overviews feature faced criticism for making bizarre recommendations based on outdated Reddit posts, such as suggesting adding Elmer’s glue to pizza, after the company secured a deal to access Reddit’s data for AI training in February 2024.
- The new features provide organizations with greater control over the information sources used by Google’s AI models, helping to reduce the occurrence of inaccurate or misleading outputs.
Empowering organizations with enterprise-ready AI: By introducing these new capabilities, Google Cloud aims to encourage more organizations to adopt its generative AI experiences for enterprise applications:
- The ability to leverage trusted third-party datasets and companies’ own corporate data ensures that AI models generate more accurate and reliable outputs tailored to each organization’s specific needs and context.
- Features like high-fidelity mode and hybrid search for Vector Search give organizations greater flexibility and control over how their AI services process and retrieve information, enabling them to build AI applications that align with their unique requirements and use cases.
Looking ahead: As Google continues to refine its Vertex AI platform and introduce new features, the company is positioning itself as a leader in enterprise-ready generative AI solutions:
- The upcoming “dynamic retrieval” feature for Grounding with Google Search will automatically determine whether information should be sourced from Google’s Gemini datasets or Google Search based on the nature of the prompts, further enhancing the adaptability and accuracy of AI-generated outputs.
- By addressing concerns about AI accuracy and providing organizations with powerful tools to customize and control their AI models, Google Cloud is setting the stage for widespread adoption of generative AI in enterprise applications across various industries.
Google touts “enterprise-ready” AI with more facts and less make-believe