Google is stepping up its efforts to improve the factual basis of enterprise AI chatbots by integrating real-world financial data from businesses and financial services.
The plans aim to reduce inaccuracies in AI-generated information and improve overall reliability, known as “hallucinations,” by substantiating answers with verified data.
The news is part of Google's plans to include third-party data sources, whose initial partners include Moody's, Thomson Reuters and ZoomInfo, in addition to the company's internal web search and data.
Google wants to make enterprise AI more objectively correct
Companies were quite late to the party; Initially, they had hoped for more secure and private systems to protect sensitive company data; However, as more companies implement AI solutions, the focus has shifted toward accuracy.
Speaking about the changes, Google Cloud CEO Thomas Kurian said Axios: In fact, you can trust the model to perform a task on your behalf because you have a basis to trust it.”
To further improve reliability, Google is also introducing a confidence score that provides a numerical indicator of how confident the AI model is in its response. Business users will also be able to direct the AI chatbot to prioritize information from specific documents or data included in a message, rather than its broader training data.
Kurian added: “We have taught the model how to ensure that when it responds, it takes what is in the input message as the main information it should pay attention to.”
Additionally, Google is expanding vector search to support hybrid searches, combining vector-based image searches with text-based keyword searches, to improve accuracy. The update is currently available in public preview.
The Google Cloud announcement, written by Burak Gokturk, vice president and general manager of cloud AI and industrial solutions, concludes: “As these technologies become even more capable, we are committed to helping companies take advantage of all the potential of ground-based generative AI in the real world.”