• Data Science
  • Anthony Macciola
  • JAN 05, 2018

Five AI assumptions and truths

Ready to learn Data Science? Browse courses like Data Science Training and Certification developed by industry thought leaders and Experfy in Harvard Innovation Lab.

Don't let these common myths trip you up as you consider your AI strategy in 2018

Artificial intelligence has ignited our imaginations over the future of work since John McCarthy first coined the term in 1955. While Hollywood often depicts AI as an ominous tool – take War Games and Terminator, and even HAL in 2001: Space Odyssey – in reality, AI has become a real and growing part of our lives, and not just in video games. Within organizations, there are real-life AI applications being deployed that automate business processes to improve the customer experience. But we are also hearing progressively more myths when it comes to business applications. Following are common myths, assumptions, and truths to know as you consider your AI strategy in 2018. 

1. An AI strategy is not one size fits all

I can’t tell you how many times I’ve met a CIO or CTO who wanted to know what our AI strategy was. I usually try to qualify their interest with a question of my own, “What does AI mean to you?” 

That question is typically followed by dead silence. The issue is that CTOs expect the same AI strategy to fit every use case. The real question that should be asked is, “What are customers actually trying to do?”

Usually, the use cases come down to basic machine learning. Customers want a system to watch and learn what their knowledge workers are doing as part of their daily routines – especially routines they are considering automating with robotic process automation (RPA). They want a system that can then start recommending various courses of action based on learned behavior. Finally, they want the knowledge worker to be able to direct the system to automate the learned behavior after getting comfortable with past recommendations.

2. AI and machine learning are not interchangeable

Machine learning is often misinterpreted for AI, as are cognitive processing, natural language processing, and deep learning. Machine learning is an application of AI whereas it gives machines access to data and lets them learn for themselves, thus enabling AI.

It’s the combination of RPA and machine learning that organizations interpret as AI. Once organizations have automated various tasks by adding a level of learned intelligence, they’ll be able to monitor and understand the impact those efforts are having on their organization via real-time process discovery and process intelligence. 

3. Algorithms are not more important than data

This is akin to the chicken-or-the-egg conundrum. Algorithms need data to work, yet algorithms enable you to do whatever you want with data. However, the design of an algorithm is guided by the data structure they are supposed to work with, so in reality, data is more important than the algorithm.

4. AI lacks an important human trait: Empathy

In reality, AI replaces mundane, repetitive and error-prone tasks so humans can focus on value-added processes that require creativity, problem solving and flexibility. However, workers carry unique characteristics like empathy and judgment that robots lack.

5. Expect a cultural and skills shift 

Organizations can count on seeing dramatic change within the company culture and employee skillset within the next three to five years. However, embracing AI needs to lead from the CEO and spread throughout the workforce in order to be successful due to the human nature’s reluctance to accept change. Having management spearhead the cultural shift will foster acceptance quicker. 

From a skills perspective, the introduction of AI into the general workplace will result in more tasks being addressed by system of record applications. For example, in the mortgage lending market, the dependency on a loan origination officer to drive the loan process will lessen due to the loan origination system being able to make intelligent decisions based on past funding behavior. This will leave only rules-based exceptions to require a loan processor’s attention. As a result, this will lighten the overall workload for loan officers, allowing them to be more responsive when an exception rises and should allow mortgage lenders to increase the productivity of their operations. 

AI will result in the automation of basic tasks performed by knowledge workers today and will have a large impact on the makeup and size of corporate workforces throughout FinTech, healthcare, transportation and logistics, and government customer/constituent engagement scenarios. 

Finally, some CIOs mistakenly believe that incorporating AI consists of a flashy robotic solution. In reality, AI and its subsets of RPA and machine learning technology are tools for digitally transforming various legacy processes of their business.

Whether it’s revitalizing the digital mailroom, streamlining the onboarding of new customers via mobile apps and smartphones, to expediting the processing of medical claims and invoices, organizations will use AI to retrain existing customers and acquire new customers. After all, digital transformation is in large part the desire to create an optimal customer experience.

With this different perspective of AI, it should, therefore, be part of a revenue generation proposal as opposed to an ROI-based purchase decision, and have more visibility at a CXO level.

Boston city bkg

Made in Boston @

The Harvard Innovation Lab

350

Matching Providers

Matching providers 2
comments powered by Disqus.