Implementing AI: When and How?

In this special guest feature, Venkat Viswanathan, Founder and Chairman of LatentView, discusses how organizations determine when artificial intelligence (AI) should be utilized to amplify human intelligence. Venkat Viswanathan is the visionary behind LatentView Analytics with more than 18 years of experience in management consulting, technology, and global IT services management. He has spent the better part of a decade helping global businesses adapt to the business challenges created by disruptive new technologies.

The use of artificial intelligence (AI) applications in business is growing, but AI and machine-learning aren’t yet an efficient use for every business task. While there are many aspects of a business which can be automated, and should be automated, tasks that require judgement, prioritization, and trade-offs still require human intelligence.

Actions that require computation or organizing large quantities of data are all better handled by machines, and applying automated solutions that use AI will wipe out other traditional business models. Organizations accumulate massive amounts of data from business operations, social interactions, and sensors. Harnessing the power of this big data to build automated solutions and provide insights will lead to sustainable competitive advantage.

With that said, how do organizations determine when AI should be utilized to amplify human intelligence?

Is AI right for your organization?

It’s understandable that the distinction may become blurry, especially as AI technologies advance. To determine the best course of action, organizations should question whether the task at hand allows for repetition, high volume, a pattern, and low cost of mistakes. The tasks that satisfy these specific criteria are starting points for AI implementation.

When considering customer requests that require decisions to be made based on empathy, AI should be avoided. On the other hand, AI can be used to provide better responses to transaction requests by efficiently collecting historical chat data. Processes at the periphery of business, in which the cost of making mistakes is low, is also a perfect opportunity to apply AI. It is processes at the core of the business that should be handled by human intelligence, as making a mistake in these high-value situations could result in substantial brand or financial damage.

Once an organization determines which tasks can be automated, it can establish how AI can be successfully implemented. To accomplish this, organizations should reference the following checklist:

Identify the problem

Begin by identifying the business problem, and ask where using AI can improve efficiency. Consider the business of a commercial airline. AI could potentially automate the next step for customers who have missed their flight by suggesting alternatives for a later one.

Identify the data source

Once the problem is identified, distinguish the data source, and collect data from relevant customer touch points. For example, an airline would use internal data from a customer’s travel history for incorporating into AI-based flight suggestions.

Develop an AI-based solution

Develop an AI-based solution to aid algorithmic decision making, and make use of neural networks and Natural Language Processing. This will help AI understand the request being made and utilize real-time data for tasks such as identifying alternative flights. It’s critical to note that AI exists to enhance human decision-making.

Implementation and action

After these steps are taken, an AI solution can then be regularly implemented. Training should follow in order for staff to effectively work with AI and understand its use and applications for appropriate situations, as well as provide human intervention when necessary.

Like with other technology innovations, we tend to underestimate how far reaching and common place AI will be in five to 10 years, while we overestimate its impact in the short term. By clarifying the implementation process, organizations can set realistic expectations on just how AI will begin to affect their business, while taking advantage of what the technology could offer.

AI is man plus machine rather than man versus machines. As machines improve and AI becomes more intelligent, the roles of human workers will change and will become more focused around tasks that require critical thinking and judgment. As a result, organizations can begin crafting higher value jobs for people that prize their creativity, empathy, judgment, and leadership abilities.

 

Sign up for the free insideAI News newsletter.