How to move past the hype to succeed with AI ?

How to push the hype to succeed with AI

Canadian CIOs are still trying to find their way through propaganda to derive real value from artificial intelligence (AI) and machine learning.

"There is a lack of understanding," said a CIO from the financial sector at the CanadianCIO Virtual Roundtable recently. "We need more substance on what AI can do for us and how it relates to our business objectives."

Many organizations are not sure where to begin with AI, acknowledged Philipp Draskovich, Information Architecture Executive, IBM Canada. "It should start with self-assessment of where they are and then plan small steps to move the analytics and AI MAT curve forward." Draskovic suggested that businesses seek "improvements in low-hanging fruits" that would produce quick results.

There was a general consensus among participants that, at the end of the day, success with analytics and AI would depend on the data.

You can't have AI without it

Draskovic stated that there is no quick fix to improve data quality. "It is not a one-time thing. It is a continuous process and two-thirds of it is people and processes."

Data quality is important to succeed with AI, but it is important to improve information architecture. "We have a saying that there is no AI without an IA," Draskovich said. Information silos have to be scrapped. Organizations should ensure that users can access and securely share data from one location.

If employees have to copy the information to share it or if ongoing improvements to the data are not tracked, you are losing time and productivity, Draskovic said. Creating copies also poses a security risk and increases storage costs.

This approach is not the same as moving data to a data warehouse. Rather, data virtualization provides an access layer to a list of data, wherever it resides. "It's a one-stop-shop for any data in the organization," Draskovic said. This resolves the biggest problem at the launch of a new project, which is knowing where to find the data. It also simplifies data governance.

Another advantage is that centralized access can put data in the hands of line-of-business experts who can become "citizen data scientists", the public sector CIO noted. They are in a good position to identify cases that support business objectives.

Are we ready to handle machines?

Many participants acknowledged that their organizations are not willing to make decisions on machines. One IT leader said, "Business leaders protest because of concerns about privacy and access to data." "There is a strong push on this front moving very quickly."

The organization should also monitor machine learning to ensure that the models they are not using become biased. This will soon be required by law, Draskovic said. An IT leader said that he allows machines to make sensitive decisions that affect people's lives.

"The bottom line is whether you trust data and governance," Draskovich said. "And it all comes down to the strength of architecture."