So much has been said over the last few days on my excellent insights on where the technology is headed, what's going on with the safeguards, and whether the investment rush is a bubble or not. And I think that today in this panel discussion, we're going to offer the best test of the best response to all of these questions by talking about the deployment of AI, which is really the test of the technology, the test of the investment, and all of that. And so I'm going to start with you, Mohammed. Get us up to speed on how you assess the pace of deployment across businesses across the world. I know no one number can sort of describe it, but what do you make of it?
You know, today, when you think about the deployment of AI globally, I think where we are at the point where there's no company that will say that I do not want to use AI. I think we're all motivated by the fact that it is real, it is here, and it will continue to evolve. What we are seeing about the pace of deployment of AI is around two dimensions. First, what are the companies that are using AI just to go for productivity improvement? Or what are the companies that are using AI just to change how they run the business, which is around the transformation of the business, but also how can they use AI to deliver more value or even grow the companies better? But fundamentally, the journey that many of the companies are going through today is to understand that there is not such a thing as AI without data, and there is also no nothing such thing as AI without governance. So a lot of the companies now embarking to the journey of making sure that they have what we call the foundations, which is do I have the right data environment, do I have the right governance, do I have the right people, the right security, etc. So it's very different from what we had a year ago. But the pace of deployment is there. What is not there yet is the ability to explain the return on investment. It's not just the yield. It's about the outcome. Right? That's what we're seeing. And different societies around the world are moving at different speed, whether you find yourself in the West, in the Global South, or in the East. But I would tell you that everybody's trying to move fast. Most importantly, it's not just moving fast. It's playing catch-up because AI keeps evolving also quite fast. Society is not implementing fast. It's also learning really fast about what AI can do.
Okay. Go through a couple of things. Maybe moving fast, but not moving deep enough. You've described this deployment process really as theater in your new book. You know, I need to have, and so I'd love for you to talk to us about whether you think they're doing the surface-level work or they're going deep enough.
I totally agree with Mohammed. It's 99% surface level, not going deep. And the reason for that is when you just ask the departments what they do, you want to build. Well, you can build lots of agents that can save a little bit of money or make a demo. But if you think about how powerful these AI models are, right, from OpenAI and other companies, these are so smart. You can say they're like IQ 300 models. If you hire an IQ 300 employee, would you let him or her be in the customer service department or driving company ROI results, or helping with new product and sales strategy, or helping the CEO run the company? So our belief is, AI transformation must begin from the top, CEO, not CIO. CIOs are important, but they make things, make sure things run. CEO is the only person whose vested interest is 100% aligned with the company. The CEO's direct reports mostly don't want to do a transformation because they are threatened by it. They're not understanding it. They think their organization may shrink. So the CEO must take the bull by the horns.
So what we've decided to do at 01.AI, that I also described in my book, is to build the agent, the ultimate agent for the CEO or the chairman or the chair or the board, whoever runs the company. And what does this agent do? It's not an ordinary agent, because this is the one agent in the company that can read every piece of data, write all the sales reports, ERP, CRM, etc. But also we add a bunch of dormant data. So every meeting. So we're doing this experiment on my own company. So we record every single meeting. It's permanently remembered, transcribed, as well as all of the email of employees, which you may not choose to do, but it's corporate email, all the corporate chats, every single sales call with the customer's permission. They'll give permission because we tell them on the service, you better. And every customer service call. So now we have this. This is so much more important because think about the meeting data. What's the most important data in your company? It's got to be the meeting data. That's where ideas are brainstormed, decisions are made, strategies are changed, and people argue. So we take all of this data. And of course, now we need to build a set of harnesses, capabilities, orchestration, and ontology. Otherwise the data will swamp the models. So I won't go into details there. So when that is done, I, as the CEO, can ask some of these questions: Of the 1,000 meetings that happened in the last week, what are some really important debates that I, as a CEO, need to be aware of? Are we going to meet our Q2 revenue target? What's at risk and how do I unblock it? Who are the best people who are running agents in my company? And are they also good leaders? So you can guess what's next. I'm going to promote them. Which of the managers are mostly spending time running meetings, managing people, and passing information? Well, guess what happens to them. So this pass AI module combined with an HR AI module is the beginning of the transformation. It's about passing the reins of the company from these fancy titles, VP, director, to people who are directly responsible for running AI and held accountable and empowered by the CEO. So in my company now, I basically have no staff meetings. I manage these devices directly, and I can be away for two months and I can know everything through my agent. So I feel this is the way to do transformation.
I like the approach. You are looking to eliminate the CEO first, as opposed to, you know, the lower-down jobs going. You're the CEO. Would you do that?
Yeah, I am trying to catch up what he's doing right now. I'll visit you.
More seriously though, I think you have been on the journey of an enterprise-wide AI deployment, and I think, if I understand correctly, more than 40% of your global staff and engineers are already working with AI. Talk us through what that journey has been like and what have been the toughest points in this deployment process?
Actually, at first, I want to talk about what I know from the kind of big picture you said about deployment. You know, we are calling that world in a kind of ecosystem, right? And we see a kind of transition from the digital world, this intelligence, into the physical world. And looking at the other participant, TDK, we are the only one who are coming from the physical world. Means that we are using AI. I know AI without data, right? You would agree. Right. And no data without sensing, right? We are one of the companies who have a wider, greater portfolio of sensors, right, in order to further derive this digital world to physical world. Right. Sensing. And also we have a kind of batteries and also electronic components and also key components for data centers, etc., etc. So we are externally focusing on supporting, contributing to this expansion, extension from digital world to the physical world. And not only business expansion contribution externally, but as you said, internally, we are also making kinds of experience. And more than two years ago, we set up a kind of AI group, organization on a global basis. And as you said, more than 40% of our team members are already utilizing this AI, and also more than 2,000 agents we have been utilizing. So looking, focusing on external business development, but also we ourselves kind of doing experiments as one of the users to further enrich the potential benefits coming from this globally growing AI ecosystem.
What has been the most challenging part of that journey, though, in terms of the internal deployment of AI?
Yeah. Two points. Encouragement. As a CEO, I keep encouraging team members to try, to challenge, or to utilize this AI. But also at the same time, governance or risk management or information security. So this governance and also encouragement, we have to go ahead with two legs. This is a kind of challenge we have to manage. The key is balance. Encouragement, challenge, but also at the same time, governance, auditing, etc., etc. We have to balance, keep making the best possible balance. I think the people would agree.
All right. Yeah. It sounds to me as if we're at very various different points on the scale, depending on which corporation or geography you talk to. And guys who really is at the top of the scale in that sense. But what are the questions you come across when you talk to enterprises? Where does the fear lie? Where are the blockages? What are their challenges in being able to move faster towards deploying AI across the organization?
I think, you know, all of what has been said here is actually the ground truth. And I think for us, I've seen the transition where, for sure, the AI decisions are being made top down. And so it's really the CEO or the board, or I would say, more like an advisory group that the CEO is creating now within the companies to actually make those decisions. And I think what we are seeing is that the next phase really for AI is really going to be not experiments, but actually driving outcomes. And so it's no more about doing pilots, but actually defining the specific problem that they want to solve, measuring how much AI has contributed in solving that problem, and then really driving customer outcomes. And so I think it's no more the stage of just deploying ChatGPT at scale, but going out, going to the board and going to the CEO and saying, what are the hardest problems that you're trying to solve? Is that moving the top line for you, or should it be something in the bottom line? And picking those three problems and going out and really working with them and solving those problems. And then I think for them to report that back to their shareholders, to their investors. And I think that is when the transformation that we bring becomes more like a scorecard and less just about, you know, I've gone and done my diffusion amongst all of my employees, because I think that was where we were last year or even earlier this year. You know, because 24 is a big example where they have an online portal where they sell and buy cars, and they deployed an OpenAI agent where it helps people onboard their cars more effectively. And so close to about a million conversation minutes is what they use our models for every month. And they are seeing that close to 12% of their lost sellers are now returning just because of this OpenAI-powered customer agent that they have. And so this is really driving business outcomes. It's no more about, you know, all of my employees have access and now I'm writing better emails, or you have the IT teams now suddenly sending you more emails, especially in the Indian context. But I would say it's really something that you can put down in your shareholder report and say, here is how I've been able to move the needle right from the top. I think the second thing is, you know, that's really at the frontier. And then you really talk about how is it changing my day-to-day life. And so I think we're seeing, of course, the whole operating model changing for someone like me as well, where there are days where I don't even open my email because ChatGPT work has become my single window to everything behind, because we have a lot of plugins. And also, you know, your ERP, your Salesforce, you know, your Kepler, all of the data points and different tabs that we had open constantly to actually allow us to do our work and make thoughtful decisions in our business on a daily basis. Now you have the single window, which has all the plugins in the back end, very simply, very easily done. And so there are days I don't open my email or any of these tabs, because ChatGPT work has become that single window for me. What I ask questions on: what is the priority? What do I need to build today? You know, how is my business performing today? Who are the five customers who I need to go out and meet? What is it that I need to speak to them about? And I think that's really changing the way we do work. And then the next phase of this is this is driving individual productivity and efficiency. How do we take that to the team level, and how do we take that to the organizational level? And which is why, or which is where I think agents working together is very important, where I have my personal agent, but how can I take it and drive that agent world at the team level, so that productivity now gets unlocked at a team level, and then of course at an organizational level. So I think that's really the challenge which we are now currently facing. But the good news is that, you know, some of the basics, which is making sure the data is in place, the context of the organization is in place, all of those things have been solved now. And so to a large degree, you know, while accountability still sits with the human, I think you have enough support coming in from the AI to help you make those decisions constantly and almost do it live. You know, today we have businesses who are finalizing the audit reports in their year-end reports and results within days, as opposed to months and weeks, because they have the data accessible to them very, very quickly. So these are the changes we are seeing in organizations. And of course, the fear is around accountability, around audit trails. Where does my data sit? How much can I control it? And I think those are the conversations we continue to have with businesses to build trust. Because, you know, we have data addressed now in India. We also have local inferencing. We've also deployed our models in ORS. So all of these things coming together, I think, are really driving significant outcomes and operational lives for companies, built on the top line and on the bottom line.
Okay. What I do want to throw up, since we've got most of your opening comments, is three or four challenges that pretty much all organizations across the world are facing, and I'd love for you to jump in with how you think leadership should be able to overcome those challenges, or what you're already seeing in companies, clients, partners that you're working with. I think the first one, to me, seems to be around costs. Costs have been coming down very rapidly, and I'm wondering how CEOs are looking at this decision-making, saying, should I jump in and start deploying now? Should I wait for costs to lower even further? What is the right time? Is there a good way to time this? The second thing with costs is the trade-off with sovereignty, which is, you know, geopolitics is at the center of the adoption conversation. And are they looking at, do I do an American model, more expensive, a Chinese model, open source, adaptable? How do I pick? Am I making the right choice geopolitically? So I'd like to start with you guys. How do you look at this entire issue of costs? And then we'll move on to next challenges.
