EdgeNetworksMedium Blogs 5 Questions Every Leader Must Ask Before Embracing AI

5 Questions Every Leader Must Ask Before Embracing AI

We are entering the age of Artificial Intelligence (AI). Gartner has forecastedthat AI-derived business value will be at a whopping $3.9 trillion by 2022, through a combination of customer experience, new revenue, and cost reduction.

With its remarkable ability to quickly process large datasets and use the information in new and interesting ways, AI brings with it an amazing range of possibilities. The popularity of AI is undoubtedly going up rapidly. 61% of businesses said they implemented AI in 2017 as per a report from Narrative Science and National Business Research Institute (NBRI).

It is safe to say that every organization can benefit immensely by using AI, whether to support strategic decision making or increase revenue and sales or provide superior customer experiences. From Netflix using AI to predict its customers’ viewing preferences to the UK news agency Press Association (PA) using robots and AI to cover and write local news to John Deere using advanced machine learning algorithms to help agricultural decision-making, there is a variety of real-life AI case studies that we are seeing.

As a leader, finding the right AI strategy for your organization is certainly an important step, considering that it could have far reaching consequences for the future of your business. But before you jump headlong into implementing AI for your organisation, there are a few critical aspects to consider which will enhance your chances of success.

Here are five questions that every leader must ask before embarking on an AI implementation in their organization.

1) Do you have a basic, functional understanding of AI?

Despite all the buzz around AI, there is often very little understanding among leadership teams as to what it really means, and how it can help their business. ‘Artificial Intelligence’ is simply a catchall term that refers to intelligence demonstrated by machines, as opposed to the ‘natural’ intelligence displayed by humans and other animals.

Just like all the other technology buzzwords of our time, such as IoT or Big Data, AI is not a magic bullet that magically cuts costs and bumps up sales. There are several layers and degrees of complexity involved when it comes to implementing AI. Therefore, it is important to have a basic understanding of how AI works, and what it can and cannot do before considering it for your organisation.

2) Have you identified the business problem?

While considering the many possibilities of what AI can deliver for your organisation, it is also important to keep the focus firmly on the business challenges that your organisation needs to address. Defining the problem statement accurately and putting metrics in place to determine the success of the implementation is the first and most important step. It is only then that you can effectively judge if AI is the right approach for you, and how you should use it for your organisation.

Also, AI is still not a mature technology. It is evolving constantly. Therefore, it is important to understand what AI can and cannot do on the ground. For example, can AI transform your entire organizational culture? Probably not, at this stage. On the other hand, if you are looking at improving productivity, then AI may be the answer. Asking the right question is important.

3) Can your current infrastructure support an AI implementation?

In most AI applications, especially those based on Machine Learning, the success of the implementation often hinges on the volume and quality of data available. The effectiveness of data depends on existing data governance processes and the current state of digitization within the organization. If the data that you are generating is incomplete or unreliable, then any machine learning output from this data is also likely to be severely flawed. So, before you rush into an AI implementation, ensure that your infrastructure is equipped to provide the right amount of data and also, the right kind of data. If not, focus on fixing the digital infrastructure before you get started on AI.

If your data is less than perfect, AI could also help in the data acquisition process to create quality data. For example, it can capture images and analyse them to generate useful data points that enhance the data quality, compared to data that has been collated manually and is subject to human errors.

4) Do you have the right team in place?

The right team can mean the difference between success and failure in any technology implementation. Does your company have right team to successfully complete an AI-based implementation? You might want to consider hiring a good AI consultant to help you put in place an AI plan. You will also need to hire data product managers and citizen data scientists to act as functional consultants to determine the right kind of AI for your company.

Training your employees with high-quality courses on AI such as this one can be considered.

5) Are you prepared to make the Investment?

However, like most technology implementations, AI needs a fairly large investment of time, money and resources, which may put considerable strain on the company bottom line, at least in the short term. Is the organisation in a position to make this investment? Also, there is always a certain risk of failure that comes with any technology implementation, and AI is no different. You need to understand and accept that there is a learning curve involved and failure is a part of the journey.

Given its incredible power to transform businesses, AI definitely needs to be a priority for any organisation that wants to stay competitive in the long run. However, taking a considered approach to AI implementation is important to ensure a smooth transition and informed decision making.


Written by
Rahul Kulhari

Rahul Kulhari is Head — Data Science at EdGE Networks. His areas of focus are deep learning and NLP. Currently, he’s working on reinforcement learning, meta learning, explainable AI and Enterprise AI.




No Comments

Leave a Reply