Winning with Artificial Intelligence

The AI honeymoon may be ending, but the partnership between human and artificial intelligence is still in early stages. Tosin Fasidi, an accomplished procurement leader, strategist and speaker, examines the key issues and shares insights on how to get it right. As the initial excitement fades, organisations are shifting their focus to delivering tangible benefits.

Transcript

SPEAKER_00 0:05

Winning with artificial intelligence will not just happen even after making the required investment in new tools. Tosin Fachi, a results-focused supply chain leader with global experience, speaks on getting it right.

SPEAKER_01 0:23

The question is not if AI would change the world. The real question is do we have the wisdom and the boldness and the humility to change along with it? For the past three years, artificial intelligence has been presented as the best technology transformation of our generation and probably for all time. We are told AI will revolutionize business, transform productivity, reduce costs, create new industries never seen before, change the way we work. And to be fair, much of this excitement was justified. For the first time, millions of people could interact with machines capable of writing, analyzing, summarizing, reasoning and generating ideas real time. The possibility seemed endless and unlimited. Companies rushed to AI strategies. Executives were falling all over themselves trying to say they are using AI. Employees were mandated and coerced to use AI. AI pilots and agents appeared everywhere. Then something happened. The question changed from what can AI do to what is the return on investment? What are the savings? What are the productivity? Why are still many projects still in better or better phase? And why are companies spending millions on AI and are still struggling to show tangible benefits? So here's the question that I'm gonna answer again. Is the AI on the moon over? My answer is perhaps it is. I don't believe the end of AI on the moon necessarily means it's bad news. You see, at the end of AI Honey Moon may really represent something far more important, and that is the maturation of AI. Now to understand why, there are a few topics we'll talk about. The first one is what happened to the AI on the moon? What's the story? Secondly, why are organizations expressing a reality check? Then thirdly, the fundamental mistakes many organizations make. And then we'll talk about what needs to change. What's the different way of looking at things? Now, everything we have starts with imagination, so is technology revolution. Every technology revolution starts with imagination. For example, when internet technology came in from the late 90s, people imagined an interconnected world. When mobile technology came, people imagined a digital world in our pockets, an entire digital world available in our pockets. Now, when generative AI exploded into our consciousness, we imagined a world where intelligence can be available on demand. Now, this excitement, this anticipation created enormous momentum, and for a good reason. Organizations began asking, how quickly can we adopt AI? How many AI tools can we purchase? How many employees are adopting AI? How many pilots can we launch? And these questions are understandable because no organization wanted to be left behind in this great thing. Nobody wanted to be the company that ignored the internet. Nobody wanted to become the company that missed the mobile technology. Certainly, nobody wants to be the company that ignored AI or missed it. So they moved quickly. They moved with a great sense of urgency. But in that rush, something happened, which happens every time. People began confusing activity with progress. A company can announce 20 AI pilots and call it innovation. A company can buy thousands of AI licenses and be termed as technologically advanced. A company can report that 70% of their employees are using AI, which is a very high adoption rate, and say they are being transformed. But here's a fundamental problem: usage is not value. Adoption is not transformation. You can be extremely busy with AI and create little value. We want to move from AI adoption to AI value. This is not an abandonment of AI. Rather, it's a reset in what we see with AI. And that reset is set in four principles I'll talk about. The first one is start with the problem and not the technology. So when you look at your your value stream or your workflow, the first set of questions you ask yourself is where's the leakage? What is the problem statement? Where why is where why do we have a problem? Why is it taking so long to go through this process? The interesting is when you go through those questions, they start discovering the root cause of the problem. Leave technology aside. Focus on the problem statement or the value statement. Then the second part, you just don't say, identify what's the problem. You need to redesign the workflow. You need to remove inefficiencies from the value chain. You need to transform the process. AI works best when you have a transformed process. AI can expose defects in the system, but to get the best out of AI, you need to transform the process first, then adopt AI later. Then thirdly, you need to measure outcomes, not usage. All this, oh, there's an 80% adoption. They are good metrics, but they're not outcomes. 80% of your employees using AI does not mean transformation. It just means that they are using it, that people are complying. Compliance doesn't mean transformation. Compliance is important on the path of transformation, but that doesn't mean transformation. Keep human judgment where it matters. Some things in our process still requires human intervention. Keep it there. Don't force AI into a space where human judgment is required. And ultimately the winning goal is human judgment we've multiplied by AI capacity. The AI on the moon began to change when executives started asking the question they should ask, which is a reasonable question, and that is what are we getting for our investment? We've spent a lot of money in this technology, we know it works, we know it's efficient, but how does it affect our bottom line? And then the conversation became more complicated. Because while individuals, and I speak for myself personally, are becoming more productive with AI, organizations struggle to turn individual productivity into enterprise value. Let's give an example. For example, an employee who can write draft a quick email that could take 10 minutes in three minutes. That's efficient, that's productive. But here's the question: did the organization lose reduce costs as a result of that? Did they increase revenue? Did they improve customer satisfaction? Did they increase capacity? Did they emit unnecessary work? Now the remaining seven minutes the employee saved. Did he or she use it for something else? More productive. So this is an important distinction. Individual productivity does not translate automatically to organizational productivity. For organization to capture value, there must be a system, it has to be at the systemic level. Efficient processes, cost must be reduced, bottleneck must be reduced, capacity must increase, revenue must grow, the risk must be managed. Because the benefit might be real but an economically invisible. Mark my words, real benefit, but economically it's invisible. And that's why organizations encounter what I can call value gap. Because somewhere between the pilot and the income statement, the value is missing. Why? Because enterprise transformation is not easy. Because this same AI technology is not acting in isolation, is dealing with legacy systems, poor quality data, requirements, restrictions, employee resistance, complex workflows, um and um regulation. So the AI model may be powerful, but if the organization is not ready, then you can't get the true value of it. I can always think about a biblical reference, which is new wine in old wineskins. Imagine a clunky 10-step process that had existed decades ago, and in this process, you have unnecessary approvals, duplicated activities, manual data entry, disconnected systems, and offs here and there. And then somebody says, Let's transform this process with AI. And all you do is you take AI, you put it in step five, and everybody rejoices. Yes, we've brought in AI to this process. But guess what has happened? You've just made a bad process faster. The process is still bad, it's just faster. And that is the challenges we have. Putting a 21st transformative technology into a 20th century process. So when you want to look back, the question is not where should we put AI into this process? The question is to look at the value chain completely and ask yourself the question: with this AI technology, what processes are to be designed? What processes should still exist? Why are five people touching the same transaction? Why are highly skilled professionals gathering data instead of analyzing data to make decisions with it? While we're asking human beings to do steps that the machines can do. And on the flip side, too, you can say, why are we asking machines to make decisions that require human judgment? Ultimately, it's not replacing everything with AI. So we'll conclude this discussion with a tag from hype to impact. So we'll say this. The second chapter is saying what value can AI create. The first chapter celebrated adoption. The second chapter would demand transformation. Here is the thing. The organizations that succeed are not necessarily the ones that spend the most money on AI, but they are the organizations that have the discipline to ask the fundamental questions. What problems are we solving? What outcome are we trying to achieve? How do we measure success? And where should human judgment remain essential? The AI on the moon may be over. But the marriage between human intelligence and artificial intelligence is just starting. And if we can approach it with discipline, with wisdom and courage, there'll be lots of value to create.