By David Roe
Please visit https://www.cmswire.com/digital-workplace/understanding-the-role-of-artificial-intelligence-in-the-digital-workplace/ for the complete article.
Organizations are Not Ready for AI
Although it is difficult to foresee what level people will be using AI to assist in their respective business challenges, it’s safe to say that AI will move from being something that’s considered advanced — today — to something that’s considered the norm five years down the road. Rather than just being tasked with doing the hardcore number crunching, AI will start to appear in other areas where it can assist professionals on the job. That may involve software acting within a certain set of parameters on behalf of a person so that the professional can spend more time allocating their daily routine to strategy over, say, doing manual data extraction.
The current state of play, though, is that most organizations are not ready for any generalized use of cognitive computing, according to Ajay Khanna, VP of marketing at Redwood City, Calif.-based Reltio, which has developed a self-learning data platform. There is quite a bit excitement about the prospects of AI and machine learning (ML), but most companies are still not ready for any serious level of cognitive computing. The key challenge is the quality of the data to attempt such an endeavor,” he said.
Machine learning requires a reliable data foundation to ensure that algorithms are acting on the right information. One challenge, not just for machine learning but for advanced analytics in general, has been the tediousness of synchronizing data models between operational applications and data sources, and downstream data warehouses and lakes that are being used as the data catalog for ML. Ensuring reliable data requires blending and correlating profile attributes across disparate siloed sources, applications, and formats. An immediate use case for ML is to help improve the data consistency, accuracy, and manageability for better data quality. ML helps with uncovering patterns and anomaly detection in data to make data stewards' jobs easy and effective. “In addition, there's still distrust in machine learning as black box magic. The initial phase of AI and machine learning must provide transparency about the rules that drive any decision, offer potential choices to the user and leave it up to users to evaluate,” he added.