AI is evolving from simple automations to cognitive agents that can perform specific tasks without human intervention, unleashing unprecedented levels of agility and performance. This technology is poised to take AI to the next level and change organizations—and life as we know it—forever. To explore that future, Mike Walsh, CEO of Tomorrow, best-selling author, and host of the Between Worlds podcast, joined The AI Forecast podcast this month. Mike talks us through AI agents from the lens of the fifth industrial revolution – which he believes stems from AI.
Here are some takeaways from Paul and Mike’s conversation.
The pandemic kickstarted the fifth industrial revolution
Paul: What's the big idea of the fifth industrial revolution and what do you see as the opportunity for our business audience?
Mike: We probably weren't on track to having a fifth industrial revolution until probably the 2030s, but then something we didn't expect happened, which was COVID-19. The world shut down. People made massive investments in digital transformation in cloud, in data infrastructure, which is appropriate for today's discussion in robotic process automation. Unknowingly, what we did was not just discover remote work in the background, we accelerated the forces that brought a fifth industrial revolution forward. And that is, I believe, powered by AI – particularly AI agents and eventually humanoid robotics. This is going to be basically a form of digital labor and the catalyst to transforming productivity in almost every industry.
The value of AI agents falls in two camps – back office and front office
Paul: Let’s dive into AI agents. Where do you think it’ll have the most impact?
Mike: Let's define what an agent is because people have different views on this. In the simplest terms—OpenAI defined an AI agent as ‘a system for taking action’.
We're moving away from the idea that AI is going to be an answer machine for all the things that we want to know, to the idea that AI is the powering spirit or the animus of systems which can proactively act on data and information and take action in commercially valuable ways. These agents have commercial intent, and that's why I think people are so excited now because a year ago people were struggling to figure out the real ROI of some of these generative AI tools. But with AI agents, it's pretty clear exactly what they can and will do.
Paul: What are some examples where this technology is being applied today?
Mike: Think of the use cases in two buckets: front of house and back of house. The most obvious areas are in the back office – business workflows, call centers, processes, even the functions of entire departments. If you can describe it, you can automate it. But if you can ask for it, you can create some sort of cognitive tool that will really accelerate the decision-making process without the boundaries of departments.
The other side of the house is something we often forget about, but it's the thing that touches us all the time, which is the customer experience. And the truth is, we will all have our own personal agents very soon. They will shop on our behalf. They will sometimes sit in on calls on our behalf. They may even go on dates on our behalf to screen partners.
If you think about that, a lot of the traffic in the future on websites and digital channels won't even be human. It'll be our agents interacting on those platforms. And this is already creating problems because if you use OpenAI's operator, which is their agent tool, there are some sites that will block you. They realize it's non-human traffic. One of the other side projects that Sam Altman is invested in is called World. You get your eyeball scanned and as a reward, it gives you some cryptocurrency. It's clear that once your identity gets locked onto the blockchain, it'll be tied eventually to a personal AI agent so that you can actually say that agent is now acting on your behalf. It's like a blue tick mark for that virtual extension of yourself. That is going to be a very real thing in the next 18 months as agents start to represent particular humans and can be authorized to make transactions and representations on their behalf.
Leaders shouldn’t be afraid of AI
Paul: You talk about the algorithmic leader, what does the future look like to the algorithmic leader? What are some of the skills they need?
Mike: To be an algorithmic leader in this new era means two things for me. It means a deep understanding of human complexity, understanding what motivates people, what drives people, what's a great customer experience – these analog qualities. But it also means thinking computationally, which is not only to know how to apply technology to a decision, but to be able to break a problem down into smaller pieces and approach the problem strategically like a poker player would. When you do that, the decision whether you get it right or wrong is not as important as the process by which you approach the decision.
That is probably the single biggest insight for any leader in this new world: it doesn't matter whether or not you've got the decision right or wrong. What matters is whether you, over time, personally (but also at scale with you and your team or your organization) are building a better system and model for evaluating and executing high-quality decisions. If you take that frame of reference, AI becomes just another tool in your kit for how you improve that decision making environment.
For the full conversation, listen to Mike’s episode on The AI Forecast here.
