The Challenges of Implementing AI: Part 1
How Artificial Intelligence (AI) is shaping the future.
Artificial Intelligence (AI) is powering change in every industry across the globe. As a result of the pandemic, the adoption of such technology has drastically accelerated. Market research from IBM found that 43% of IT professionals worldwide reported an accelerated rollout in their organisations since the start of 2020.
While these rapid advances have unlocked new possibilities for innovation, significant challenges remain for companies seeking to develop a holistic and scalable AI strategy. Our recent roundtable discussions with executives from industries including Banking, Manufacturing and Automotive to name a few, identified the below as the six biggest challenges organisations were facing when adopting AI:
- Recruiting the right skill set and keeping top talent.
- Complying with international data security regulations.
- Getting and using data to its fullest.
- Using AI to supplement business strategy as a whole.
- The pros and cons of on-prem vs cloud security.
- Implementing natural language processing and conversational AI.
So, this 3-part blog series will explore those challenges, and what senior executives are doing about them. This first part will explore recruiting and retaining the right talent and implementing natural language processing (NLP).
Recruiting and Retaining the Right Talent
“The main concern for us is to do with expertise, the question of who’s around to help us, as well as the migration of data, how to treat it, how to make use of it.”
– Schneider Electric
When recruiting, even the experts can find it daunting to know what to look for and what questions to ask. “This stuff is complex! Finding people with the right skill set is a challenge,” shares a senior executive from Clarion Housing Group. A delegate from Xerox emphasizes this is an area she feels they are doing well in. “It’s exactly the reason why I’ve just completed a postgraduate in AI.” She shares there are much shorter, and more manageable courses you can do. “Although I have an engineering background, I know that AI is a completely different space from what I studied. I needed to know the details on how to recruit and make sure that you have the right people, and that I asked the right questions during recruitment. So, I think it’s just on us to just keep training ourselves.”
It can be easy to fall into the mindset that you need to recruit someone who already knows everything about the technology you already have or want to move to…. but the tools are moving so quickly that it is unlikely many candidates will know it all. One executive from NVIDIA says that if you are working on innovative technology, it becomes extremely hard to find people with specific experience in that, so you must go into recruitment with the mindset of ‘we will figure this out together.’ Then, upon finding the right talent she advises you need to “support them in working on cutting edge stuff is and give them the means to do their life’s work.”
Implementing Natural Language Processing and Conversational AI
“One of the fastest moving domains in AI is now NLP.”
-NVIDIA
A theme that comes up on roundtable discussions: Natural language processing (NLP) and Conversational AI. NLP is a technology that leverages computers and software to derive meaning from human language — written or spoken. Conversational AI is the application of machine learning to develop language-based apps that allow humans to interact naturally with devices, machines, and computers using speech.
There is so much evidence to show that organisations who have introduced NLP techniques and conversational AI into their customer interactions end up reducing churn, lowering cost, and improving the experience. However, those organizations working in Europe are missing locally developed NLP models. An executive from NVIDIA shares, “there is an unofficial leader board for NLP models, and if you look at the leaders and the languages, you see that there is not a central European company, nor a Central Europe language in there.”
So, while a lot of work is still to be done with the improvement and implementation of NLP worldwide, AI-driven services in speech and language present a revolutionary path for personalised natural conversation, but they face strict accuracy and latency requirements for real-time interactivity.
For more information on all of these topics around AI, join us at our CIO summit in June, and look out for the upcoming part 2 and 3 of this series. Next up, we will look at managing data to its fullest potential and data security regulations in AI.