California Apparel News

Announcing new capabilities for its Watson business artificial-intelligence technology, Armonk, N.Y.-headquartered IBM revealed fresh approaches that it hopes will build trust in data and AI. According to its October 2020 survey, “Scaling AI, Not Risks: Removing Trust as a Barrier to AI Adoption,” IBM discovered that 84 percent of AI professionals surveyed feel consumers are more inclined to use services from companies whose data and AI models are built, managed and utilized through transparent, ethical frameworks. Through IBM’s updated Watson Studio, IBM OpenPages with Watson and IBM Planning Analytics with Watson, global brands are adhering to workflows that afford greater trust.

“AI is only as useful as your trust in it. Especially in business and when the stakes are high,” explained Daniel Hernandez, general manager of data and AI at IBM. “IBM Watson continues to deliver critical new capabilities to help enterprises build trust into every step of the AI lifecycle so business leaders can confidently operationalize AI across the hybrid cloud.”

The importance of this Watson advancement lies in its ability to help companies reduce the negative impacts of data or AI bias. During its survey, IBM discovered that 82 percent of AI professionals who were surveyed felt that their organizations were affected negatively due to problems with data or AI models.

To help companies avoid these issues, IBM OpenPages with Watson now features a new data-privacy-management module that shows a real-time view of how private data is being used throughout an organization. This feature also aids in report automation of personally identifiable information, which increases accuracy while reducing time spent during an audit. IBM Planning Analytics with Watson features an updated statistical-details page that affords more transparent and clear facts regarding the processes through which forecasts are generated.

Within IBM Watson Studio, the technology company reported the implementation of new federated learning capabilities as a tech preview. This feature helps businesses implement techniques in machine learning in order to apply it under circumstances during which data movement is not recommended due to circumstances such as data privacy, secrecy, regulatory compliance or the size of the data. New federated learning capabilities within IBM Watson Studio allow training of AI models on previously siloed data sources. IBM Watson Studio’s new time-series capabilities that help businesses automate, analyze and forecast time-series data, such as those in finance, manufacturing and retail. This advancement will help predict future values of time series using past data or features, achieving accuracy for a number of univariate datasets such as information from phone-call data logs, weather data, travel times, retail sales and production volume.