Developments in artificial intelligence (AI) are moving so fast that it can be difficult to keep up and focus on the most important issues at the same time. At our latest Think Ahead event, we brought together three expert panelists to shed light on this rapidly changing technology. Nikos Sava is Professor of Management Science and Operations and Academic Director of the Data Science and AI Initiative at London Business School. Lilia Christofi, EMEA Financial Services Data, AI and Technology Partner at PwC Consulting. Iryna Tsyganok EMBALS2023, CEO and co-founder of AI consulting firm Irysan. Their discussion was led by journalist and author Stephen Stern.
Nikos was clear about how he sees the central challenges at the moment. “Technology is evolving very quickly, which for me is both exciting and a little scary,” he said. “And what I find a little scary is that there is a gap opening between the capabilities of technology, on the one hand, and our ability to effectively deploy it into complex, human-based systems. In the short term, I think it’s about how to best prepare and evolve our organizations and skill sets to take full advantage of technology. But I also worry about the long-term implications: What will society look like in a world where AI is abundant?”
Irina Tsyganok brought the discipline of systems thinking to the discussion, saying, “AI will really amplify good systems, amplify bad systems, and force organizations to rethink the status quo. Is that enough? Are you ready to embrace AI?”
Lilia Christofi’s experience working with a wide range of financial services clients across different markets in EMEA has led her to ask practical questions about AI implementation. “How do we make sure we create a humanizing experience for consumers that is actually beneficial, ethical, and helps them grow their wealth? This is a huge challenge.”
The first of two audience surveys revealed that 38% of viewers were “experimenting with pilots and proofs of concept.” In this anonymous online poll, only 19% said they didn’t know where to start. Perhaps the high quality and expertise of the audience explains why this answer has a low score.
When it comes to commercializing AI, Lilia felt that we are moving from proof of concept to now “getting really into the core processes and starting to build scalable solutions.” While progress won’t come easily, “the ambition is definitely greater than last year, and behind it there is investment from a strategic, top-down perspective to drive AI transformation.”
Nikos pointed to the challenges of regulating this new technology. He said: “To be honest, we don’t really know how to address these issues because technology is evolving so quickly. [regulatory] problem. Therefore, more research is needed on this. We need an open mindset. And playbooks are being rewritten as quickly as technology is being developed. ”
The discussion also called for balance and proportion. Yes, AI “hallucinations” are a problem, but humans can also get things wrong, Lilia said. However, Irina warned that “the impact of AI is much greater than the impact of any single person.”
There has been much discussion about the need for “human involvement” in decision-making, but Lilia forcefully argued that this is the wrong approach. Think about contact centers. “There are thousands of interactions going on. You can’t involve humans in every single decision. It’s impossible to achieve the desired productivity gains and keep humans in the loop. So you have to consider that humans are on top of the loop, and that means managing through escalations and exceptions. It’s really important to have well-defined boundaries and be able to control them.”
How should we think about AI? Nikos said this should not be thought of as a computer-like system that, given the right input, will always give the right answer, but as a junior colleague who, while enthusiastic and intelligent, often lacks context and does not necessarily have the judgment that senior leaders have over years of experience. “Therefore, careful guidance is required.”
Irina flagged another issue worth considering. “To me, one of the problems with AI is that it’s very persuasive. And it’s about right. Often that ‘about right’ is good enough, but sometimes it’s not.”
Nikos said organizations would make a mistake if they expected to see big improvements just by giving their staff access to AI. “It’s not like we’re going to license Copilot to our legal department, our customer service department, our sales department and expect a huge change in productivity because of the introduction of AI,” he said. “It comes from thinking end-to-end about how value is created and trying to redesign how we create this value now that we have this technology available.”
Lilia agreed. “We have this new capability that didn’t exist before. Let’s not just plug it into an existing process to make it a little bit better or a little bit faster. Let’s think about what else we can do now that this technology exists.”
Nikos continued, “Assuming we know how we want AI to behave, there’s a second problem, and that’s the problem of coordination. That’s how do we ensure that artificial intelligence does what we want it to do? I don’t think we’re investing as much as we should in safety measures and understanding how to regulate, control and coordinate the technology. That should worry all of us.”
A second audience found that 43% of the audience had difficulty translating their AI ambitions into real operational change. But for a significant 33%, redesigning jobs, skills and responsibilities was an issue as well.
So what will happen to humanity in this technological age? Irina felt that humans would continue to be needed. “I don’t think the human element will ever go away,” she says. “If anything, I think it will become more pronounced and more of the responsibility will be shifted to humans to make the right decisions, because every decision we make will lead to a larger impact when made using AI.”
Finally, Nikos urged us to distinguish between the specific tasks that AI can perform and the jobs that will continue to be performed by humans. “As AI automates some tasks, it increases the value of the remaining tasks that still need to be performed,” he said. “That could require even more experts. So while it is not a foregone conclusion that AI will make jobs obsolete, it is a foregone conclusion that AI will change the way work is done.”
The discussion addressed a central challenge: the gap between how quickly AI capabilities are advancing and how slowly our educational institutions are adapting. While revolutionary AI is on the horizon and artificial general intelligence (AGI) already exists, some argue that AI governance is not yet in place. With many companies still struggling to create basic AI tools that deliver meaningful benefits, how do organizations respond? The honest thing to do is to build your organization’s capabilities quickly and admit that you don’t have all the answers as a first step.
