By 2019 application functions based on AI will be pervasive in 90% of the organizations: Van L. Baker, Research VP, Gartner

Artificial Intelligence has significant benefits because it can deliver breakthroughs in business process and redefine the ways of doing business. In an interview with ETCIO.COM, Van L. Baker, Research VP, Gartner discusses the state of AI adoption across the enterprise landscape and the best practices organizations need to embrace in their journey to be an AI-powered business.

Artificial Intelligence and Machine learning are the buzzwords and often seem to be used interchangeably. This perception often leads to confusion. Can you explain the difference?
Artificial Intelligence is about human intelligence exhibited by machines. Machine learning is an approach to achieve Artificial Intelligence. It uses algorithms to parse data, learn from it and then make a determination or prediction about something. Deep learning is a technique of implementing machine learning by using multilayer neural networks.

Which are the areas where AI can be brought to bear?
The business applications that most organizations have already have AI. Business applications of Microsoft, Salesforce, SAP, Workday already have AI deployment within them. So organizations don’t have to build it themselves.


I recommend AI for activities that can differentiate the enterprise rather than do cost reduction. In terms implementations, the low hanging fruits are the conversational AI/ Virtual assistants. We’re seeing very widespread interest in that area. We expect to see broad scale activity and PoC in this area. There will be an active adoption of the conversational interface because of the value proposition for the users in very compelling

Customer engagement applications, customer center service, and support, digital marketing platforms are some of the key areas where AI is being deployed. Some forward-looking enterprises are also adopting in for cybersecurity, financial management systems, manufacturing, and operations.

What are the key challenges to the adoption of AI within organizations?
Skill gap, use case identification, strategy, and funding are key challenges to be overcome for organizations to adopt AI.

According to a recent survey conducted by Gartner, 54 % of the organizations deal with the lack of necessary staff skills, , 37 % are struggling to define their AI strategy, 35 % are scrambling to identify the right use cases for AI, 35 % are finding it difficult to get funding for AI initiatives, 30% are dealing with security or privacy concerns. There are also issues around the complexity of integrating AI with our existing infrastructure and determining how to measure value from using AI.

Access to data is one of the biggest impediments to the adoption of AI. How can organizations deal with it?
That’s right. Without the access to right data, AI projects are doomed to failure. There have been many high profile failures of AI projects.

Enterprises need to be very focused on data management, data cleansing, and data integrity. A lot of the organizations have data residing in multiple repositories. The source, origin, and authenticity of the data are debatable. So they have to find the data, cleanse the data and structure it in the right way.

How can enterprises prepare to take advantage of artificial intelligence?
Enterprises must make the use of AI features of packaged business applications pervasive in their organizations. They should start development of conversational agents as the onramp for use of AI services.

Organizations should test hosted AI services to determine the efficiency of standard models and the amount of data needed to add domain specific expertise. In addition, they should look to gradually incorporate deep learning and machine learning skills whether via acquisition or outsourcing

They should assess the skills on their teams on ML/AI capabilities and establish training programs for those that have the capability to learn.