IIA guides data and analytics leaders and their teams in the transition to enterprise-wide advanced analytics and AI. Our passion is helping data and analytics organizations at complex companies navigate the human hurdles in implementing data and analytics initiatives, challenges like stakeholder engagement, business alignment, and user adoption. IIA’s Research and Advisory Network (RAN) is where clients plan and execute key initiatives through 1:1 conversation with a diverse network of high-performing practitioners and industry experts. IIA’s RAN professionals help facilitate these conversations and consult clients on key challenges. We like to think of this service as your sparring partner for analytics maturity. We’ve been nurturing this exclusive network since 2016 and currently have 150+ amazing members. This article is part of the series, “Open Source: IIA Experts in Conversation,” where we get inside the minds of a diverse range of experts available to RAN clients.
In conversation with Mike Congdon, Head of Data and Analytics, Southern Cross Healthcare.
In the ever-evolving landscape of data and analytics, staying ahead of the curve has become synonymous with success. While data literacy has long been a cornerstone, a new star is rising on the horizon—AI literacy. As machine learning and artificial intelligence applications become more pervasive in businesses, data and analytics leaders find themselves at the forefront of fostering AI literacy within their organizations. To shed light on this emerging discipline and its impact, we turn to the insights of IIA Expert and seasoned data and analytics leader, Mike Congdon.
In this conversation, Mike shares his perspective on defining AI literacy, strategies for promoting it, debunking AI myths, and glimpses into the future of this vital skill set. The conversation has been condensed and edited for clarity. Jason Larson, director of content at IIA, conducted the interview.
1. Mike, first things first. Welcome to IIA’s Expert Network. Your leadership of greenfield data and analytics implementation is a tremendous resource for our Research and Advisory clients who are in the beginning stages of building out the data and analytics function. I’m sure you’ll have many lively discussions. In your current and past roles, I know you’ve had a lot success developing data literacy programs for the organization, and I think we’d all agree that data literacy should be a key element in any organization’s data strategy playbook. With the popularization of AI over the past year, it feels like we’re entering a new phase of data and analytics literacy, let’s call it “AI literacy.” What’s your take on this evolution? How would you define AI literacy?
Thanks Jason, and I’m thrilled to join IIA’s Expert Network. I look forward to getting to know some of your clients and helping them along their journey.
How would I define AI literacy? This is a good question, and every data and analytics leader should be wrapping their arms around this. I ascribe to the fact that we’re in the Fourth Industrial Revolution and being AI literate is an increasingly important aspect of modern society. AI literacy is all about the ability of people to understand—and effectively engage—with AI technologies. In my opinion, AI literacy involves developing a foundational knowledge across three primary aspects of AI:
- Applications. Concerns the practical use of AI.
- Principles. Concerns the ethical guidelines for AI development and use.
- Implications. Concerns the wider consequences that arise from the application of AI.
With this knowledge people can have more informed conversations around the benefits of AI solutions versus potential risks.
That framework for developing foundational AI knowledge makes a lot of sense. So, what might be some key lessons from your experience in promoting AI literacy within organizations?
My top five lessons would be:
- Executive support. With executive-level sponsorship championing AI related initiatives it adds more validity and importance to related learning opportunities, like literacy. All you need is one influential exec to champion AI and it will pay great dividends.
- Promotion and involvement. Creating discussion groups and forums, both in person and online, facilitates collaboration and the sharing of ideas, and generally is a great way to bring about more clarity and understanding of an organization’s AI program, and just AI in general.
- Know your audience. Make sure that the literacy program is flexible enough, in terms of how it is delivered, and topics covered, so that it can be understood by all participants.
- Emphasize benefits. To some, AI is still a little scary due to common misconceptions that it automatically means job losses, and then there are other more alarming perceived implications (think Terminator). Yes, it’s important to talk about the risks when promoting AI literacy (there’s risk with any technology), however it’s most important to discuss the great benefits like new business opportunities, better decision making, and increased efficiencies.
- Make it real. Showcasing real-life AI technologies is a great way to address the “so what” question by providing tangible examples of AI adding value.
Perhaps related to the “know your audience” lesson, what strategies have you found most effective in getting non-technical staff excited and engaged with AI concepts and technologies?
Creating an algorithm that automates a manual task that an employee performs for a couple of hours per day, which allows that employee to focus on the more value adding (and interesting) aspects of their role, is a very effective way to capture non-technical people’s attention. Then periodically showcasing work, another lesson I mentioned, to a wider business audience, further builds engagement.
The two most important things for me would be communicating in layman’s terms so that people don’t just switch off because they do not understand, and as I said earlier, making it real really makes the difference. It instantly bridges the gap between theory and reality. It really is much more effective to “show” people AI in action, rather than just “tell” people about it.
How important is it to build an AI development/delivery team that includes both technical and non-technical members in fostering AI literacy?
Very important. And not just for AI literacy, but also for analytics in general, for the same reasons. From my experience the most important influencer of success in analytics is not the kit that is used, the processes that underpin the service, or even the outputs (descriptive, predictive, AI, and so on), it’s the people. The providers and consumers. The glue that holds this together is usage, underpinned by effective communication. This is why roles such as Analytics Translator and Data Journalist/Storyteller are increasing in prominence. These types of roles can bridge the gap between the technical experts within an analytics team and their usually non-technical business stakeholders. This guarantees business requirements, and resulting insights, are fully understood.
Ok, let’s shift gears a bit and discuss challenges you’ve faced when educating people on AI, especially for those employees who may not have a background in data science or machine learning?
The best way to approach this from my experience is to avoid, as much as possible, technical terms/jargon, acronyms, and focus on the “what” and “why,” as opposed to the “how.” To give an example, if I am discussing a potential AI solution that identifies abnormalities in medical images with a non-technical employee, I’m going to be focusing on benefits, efficiencies, outcomes, risks, costs, those sorts of things.
The fact that this can be delivered via a convolutional neural network, and the finer details on how this works, I’m not going to go into all that unless specifically asked, and then I would opt for a layman’s terms description of how this works, something like: medical image recognition is like a smart computer program that's great at understanding pictures. By breaking down images into smaller pieces, it learns to recognize basic features and shapes, and then combines that knowledge to understand more complex patterns, which helps it to differentiate objects and recognize like objects. That was on the fly, we could probably simplify this description even more.
Going into the detail of the “how” can often result in confusion and frustration for those who are not technical minded, and can really derail a conversation, especially one with a senior leader.
Research and Advisory Network
Access the leading network of independent analytics expertise, allowing you to apply real practitioner insights against initiatives, projects, and problems.
You mentioned how important executive sponsorship is for AI literacy. The general role of leadership in AI literacy seems huge. Talk more about the role leadership plays in supporting AI efforts.
Without a doubt, the exec has a critical role to play in supporting an AI program. Here’s a few key benefits off the top of my head:
- A clear and compelling AI vision and strategy that underpins the wider business strategy need to be championed by one or more exec sponsors. This is the greatest benefit so everyone understands the significance of the program in the broader context.
- An exec-level steering committee should be formed that oversees program activity and progress, and ensures that everyone is aligned, informed, and supportive of the program’s objectives and outcomes.
- Make sure that the program receives sufficient human, financial, and technology resources. This includes staffing the right skills, sufficient budget, and access to necessary tools and technology. Each of these areas can stymie program implementation if not cared for appropriately.
- Address any organizational barriers and bureaucracy that could impede program workflow. This could involve process or policy changes to pave the way for deliverables.
- Encourage cross-functional communication, and collaboration, among the different teams involved in the program, if there is more than one functional area.
In your opinion, what are the fundamental AI concepts and skills that every employee, regardless of their role, should be familiar with?
I would go back to the three primary aspects of AI I mentioned at the top of conversation. The various applications of AI technologies, the principles that should be utilized to prevent unintended consequences (among other things), and the implications of developing and deploying AI technologies.
Aside from this, a fundamental understanding of AI, Machine Learning (ML), what an algorithm is, and how the three relate to each other, would be beneficial. The way I would describe these concepts in layman’s terms is: AI is simply the concept of making machines smart so they can do some of the tasks that a human can do. ML is a subset of AI that focuses on giving machines the ability to learn from data, and algorithms are the instructions that guide machines to perform AI and ML tasks. All three often work together.
Now the perennial question about measurement…how do you measure the success of AI literacy initiatives, and what metrics or indicators can be used to gauge progress?
There are various methods to do this, some more involved than others. It’s really dependent on the level of sophistication or maturity of your AI literacy program.
Feedback and surveys are a pretty simple way to assess the satisfaction of participants in terms of perceived value, and confidence in navigating AI related topics after completing the literacy initiative. Assessment tests can also be performed to measure participants’ understanding of fundamental AI concepts, terminology, and applications, before and after the program.
Ongoing engagement and participation of people in AI related discussions, events, or communities, can also be assessed, which is a good indicator of sustained interest and application of AI knowledge.
For the more technically minded participants, success can be measured by assessing participants’ ability to apply AI concepts in practical scenarios like problem-solving exercises and hands-on projects.
Have you encountered any misconceptions or myths about AI that you needed to address as part of your AI literacy efforts?
Yes, definitely. As I said earlier, to some, AI is still a little scary because of common misconceptions. Some of the main ones I’ve heard from people are:
- That AI will surpass human intelligence and decide to enslave humanity! This rather alarming myth has been perpetuated by various science fiction movies over the years. But in reality, AI tools are created and controlled by humans and their actions are limited to the purpose for which the AI solution was created for.
- That AI will lead to mass unemployment by automating all jobs previously performed by humans. Now, while AI can certainly automate some tasks, in my view it’s more likely that AI will augment the capabilities of humans by transforming existing roles, more often than just completely eliminating them, and creating new roles.
- That AI possesses human like cognitive abilities and can think, reason, and learn in the same manner that humans do. This is not the case. AI systems are simply algorithms for the most part and many operate based on statistical correlations and patterns in data. AI lacks true understanding and self-awareness.
- That AI has just popped up within the last couple of years. This misconception is driven largely by the emergence of ChatGPT. In reality AI has been around for decades in one form or another.
Looking ahead, what trends do you see in AI literacy, and how do you plan to keep your organization's workforce updated and informed about the latest AI advancements?
I think the discipline of AI literacy will follow the same trajectory as data literacy. As the application of AI becomes even more pervasive, and increasingly included in organization’s technology strategies, so will AI literacy initiatives.
Keeping abreast of the latest trends in AI is something that all analytics leaders should be doing, and there are various methods to do so such as following influential figures and organizations (such as IIA) online, online communities, newsletters, webinars, and courses, to name a few.
To share this knowledge with their organizations there are various options such as implementing periodic training sessions and workshops focusing on AI advancements, inviting industry experts to share insights and practical applications, establishing communication channels to regularly pass on news and articles, like Intranet sites or newsletters or existing internal analytics communities and forums.
These are great ideas to stay up on the field and support this idea of ongoing AI education within the enterprise. Thanks for your time, Mike, and good luck on your future AI literacy efforts!