5 Key Questions to Consider before Investing in Artificial Intelligence

Source: Adobe Stock Photos. This post originally appeared on LinkedIn.com here.

As a business leader in Artificial Intelligence (AI), I usually come across two kinds of professionals; devotees and doomsayers. Devotees blindly follow AI; they believe it can be used to solve any business problem, even world hunger and peace! Doomsayers, including prominent leaders such as Elon Musk, vehemently claim that robots are going to take over humanity. As you can imagine, the truth lies somewhere in between both these extreme points of view. Of course, AI has shown enormous potential for disrupting traditional business models, creating tremendous business value, and paving the path for data-driven economies. However, AI is also at the top of its ‘hype’ cycle with extremely inflated expectations. Honestly, as a leader responsible for developing enterprise-wide AI strategy, I often struggle to differentiate real substance from fluff. I have taken a number of wrong turns during my own AI journey and learnt a few painful lessons along the way. So, here are 5 questions you should ask yourself before making the next investment in AI.

1) Where is the proof?

AI Business Development professionals have perfected their craft. You will be pitched amazing ideas dressed up as a once-in-life opportunity packaged in glossy PowerPoint decks and backed by impressive client lists. Trust me; it is OK for you to be a cynic. To determine if there is any real substance/value, you will have to dive deep into their Machine Learning algorithms. Ask them to provide supporting research papers from peer-reviewed journals. Test and learn about new AI capabilities in your own lab environment by running champion/challenger tests and proof-of-concepts studies. This is very tedious, time-consuming, and resource-intensive investigative work. But it will lay a strong foundation and protect your investments in the future.

2) Can we quantify the value?

To create value, you need to make your products/services better, faster, or cheaper. It is as simple as that! Identify the actual source of value from AI. Do an old-school cash-flow valuation exercise and get clear buy-in from all stakeholders on key assumptions. Your biggest regret would be not going back to the original business case to test the validity of initial assumptions. Remember that it is OK to make a mistake but irresponsible to repeat it.

3) How would AI benefit the end customer?

AI will create tremendous business value but are you willing to share it with your customers? Use AI capabilities for identifying the root cause of friction points, tracking consumer sentiments, and building long-term brand loyalty. Without a strong connection with customer experience, AI capability will likely be limited within the technology domain and struggle to align itself with enterprise business goals. Measure the impact of new AI capabilities on customer experience metrics such as complaint rate, net promoter score (NPS), or customer effort score. Such measurements would help evaluate the real impact of AI adoption for your customers.

4) Can you execute it?

Finally, the rubber hits the road! Use the golden triangle of people, process, and technology to determine if you are ready for prime time.

“Vision without execution is just hallucination”
- Thomas Edison

First comes the people. Good AI talent is worth its weight in gold these days. However, most of the top talent is already employed or being aggressively recruited by FANG (Facebook, Amazon, Netflix and Google). Which means, the rest of the mortals like us will have to develop our own human capital. Invest in Data Science and Machine Learning education programs to develop the internal talent pool. Secondly; establish a comprehensive process for getting executive buy-in, harnessing innovation-driven culture, and developing AI governance. Lastly, technology is evolving at the speed of light so fostering a futuristic attitude is of utmost importance. Deploy Big Data frameworks and powerful yet user-friendly Machine Learning toolkits to enable the creation and implementation of AI-enabled products.

5) Is it the right time?

Even if you answer with a resounding ‘yes’ to all the questions above, it is often difficult to judge whether this is the right time to jump aboard the AI hype train. Here is a simple suggestion. If AI can enhance your organization’s core competencies and help develop inimitable long-term competitive advantage, then go ahead and make an investment in AI right now. If AI can only enhance few peripheral competencies, the return-on-investment (ROI) on new AI ventures would most likely be underwhelming. In such cases, it might be prudent to wait for the AI hype cycle to pass and reach maturity for mainstream business adoption.

Hopefully answering these 5 simple common-sense questions will immunize you against the ‘Shiny New Thing’ syndrome and guide you along the holy path for building innovative AI capabilities for your business. Do you have any suggestions to expand this list? Please use the comments section below and let the conversation begin!

This post originally appeared on LinkedIn.com here.