Mission

Search

Differentiating AI from IT with Dan Faggella, CEO Emerj

Play episode

Or listen in your favorite podcast app

Apple Podcasts  /  Google Podcasts Spotify

As artificial intelligence becomes more prevalent in businesses, leaders in the C-suite are trying to figure out when and how to deploy the technology. Many questions will come up within organizations as they deploy AI, and luckily, Daniel Faggella is here to help answer them.

Daniel is the CEO and Head of Research at Emerj Artificial Intelligence Research, where he and his team map the capabilities of AI and hone in on where to deploy AI to get the greatest ROI. In this episode of IT Visionaries, Daniel explains the work that goes into making those kinds of projections and he discusses how the AI field is still growing and changing. 

Key Takeaways:

  • Understanding where and how to deploy AI is critical, and will be even more important in the coming few years.
  • Emerj thinks about AI advantage across two spectrums: critical capabilities and data dominance.
  • One of the big hurdles in AI adoption is educating people about the difference between AI and IT. Emerj has identified three main points they use to educate clients on the differences.

How Emerj is Mapping the AI Space 

Emerj is focused on mapping the capability space of AI to help leaders make powerful artificial intelligence strategies. They track the total AI use cases on the inventory side as well as the enterprise side to get a sense of what the company’s deployments are and how they allocate funds. After that information has been mapped, Emerj has an understanding of which strategies to deploy and where to deploy them.

For example, a banking client of Emerj’s is getting started with AI and has implemented it in just a few areas. They are now identifying what critical capabilities are valuable and where they can apply those capabilities elsewhere in the company.

“What would be the critical capabilities we really want to build and focus on? And what would be the key pillars of AI deployments that we’d really want to nail first? Because as it turns out, it’s extremely hard to deploy these systems in existing enterprises. And it’s also somewhat challenging to vet for the right vendors that are actually going to have a chance of delivering results.”

What is an AI Advantage? 

Emerj is focused on the ROI of AI and they think of AI in two spectrums. One is critical capabilities. There is a whole new set of skills that needs to emerge and there has to be a new understanding of data, resources, and leadership. Being able to build that foundation makes a company more nimble moving forward. An example of a critical capability is the idea of building a new IT foundation that enables AI and allows you to quickly jump on other opportunities to deploy AI when they arise. The second key advantage point that Emerj looks at is data dominance. Data dominance is the idea that you need to build a critical mass of data in a particular, narrow domain to leverage an advantage using AI.

“Artificial intelligence is clearly not the only way to build an advantage in business. You can have better operations in different ways. You can have leadership. We’ve just decided we’re very focused on one thing. And that’s the ROI of artificial intelligence.”

“One element of advantage is if we can build the new kind of foundation that’s going to enable AI, then we’ll be able to… nimbly move to the applications that have high ROI.”

“Data dominance is the golden dream of the venture capitalists.” 

 Who is Deploying and Not Deploying AI Well?

Companies that are born digitally-native have an advantage when it comes to deploying AI. But everyone in the industry still has quite a tough road ahead. Some of the spaces that appear to be moving more efficiently are entities like e-commerce and online media. Neither have decades worth of experience, but they are adopting quickly because of the need. One of the key trends has been that big, “stodgy” companies that are not traditionally customer-facing are having a hard time deploying AI.

“There are going to be more challenges with applying AI to sales, that’s for sure. The question would be if we were looking for an AI opportunity, and that’s normally when people come to us…then we would want to look at, well where is the data and where is the potential data advantage?”

Operational Efficiency

With all the uncertainty surrounding the current pandemic, one thing Emerj is really having to take a look at with their clients is operational efficiency, because a lot of their current strategies are having to be rethought. Once everyone comes out of this situation, Daniel believes companies are going to remain focused on operational efficiency to become leaner and meaner.

“What are the elements of what we’re doing within our company that honestly could be smoother, could be faster, could be more automated, or could be augmented so that they’re speedier and less costly for us than they were before?”

Avoiding AI Buzzwords

A big problem in the AI space is the use of buzzwords — people using terms or phrases in order to garner some type of attention. In order to build an AI strategy in a high-ROI way, leadership needs to have a fundamental grasp of what everyone is working on within the AI landscape. The idea that AI is just another form of I T— something that can just be plugged into – cannot be the mindset of leadership.

“We could call it an uncomfortable reality of artificial intelligence where in order to build a technology strategy that leverages AI in an advantageous and a high-ROI way, we do need leadership that has some fundamental grasp of what the heck we’re working with here.”

“AI is such a new domain. It’s such a new world. There are so many new facets of what we have to grasp to get it to be unlocked.”

Educating Leadership

Emerj is helping educate leadership by focusing on three areas. The first area is how machine learning works broadly, which entails how machine learning strategies are trained. The second is basic terms and components. And lastly, a representative set of use cases.

“Even just knowing a fistful of use cases for an insurance executive who otherwise kind of thinks it’s like IT will help to unlock parts of the brain to get their ideas flowing in new ways where they can actually contribute to that conversation and build a strategy that’s realistic.”

Misunderstood Aspects of AI

You have to keep beating the drum that AI is not IT. For IT, integration means finding APIs, connections, and doing all the heavy lifting to plug it in. With machine learning and AI, the focus is on your data assets.  You have to integrate and experiment to see if there is even a fit between the data and the program to begin with. There are very few AI projects that possess some kind of certainty. So for AI, you need to have a more R-and-D type of mentality.

“AI is going to always be an experiment, and sometimes experiments fail.” 

“For a lot of enterprise use cases, we really cannot be making bold claims upfront. We have to be able to take initial steps and that’s something a lot of companies do not want to stomach.”

Leading Through a Crisis 

Technology priorities are changing. Leading through a crisis involves more than just knowing how to use AI; it involves identifying the problems that are most critical to solve right now and figuring out what technology can help solve them. Companies will inevitably be focused on having resilient operations when things rebalance in the future. Companies least far ahead in their digital transformation will be the hardest hit. 

“There is a pivot toward applications with ease of deployment, measurable ROI for risk reduction, things that we can get going relatively swiftly. So we’re seeing any innovation effort that’s still left being pivoted toward those kinds of priorities.”

Mentions:

Menu

Episode 165