Live: Proof over promise – Insights on real-world AI adoption

本站添加: 2026-05-07

[00:00:00]uh how AI will transform industry at CL we're looking at the value creation because industrial AI is not like consumer AI you you ask a question and the AI provide answer to you it's not just simple query and answer for industrial AI it's a highly complex there's a many application scenario there are many variables some are controllable some are not controllable uh with such a complex decision making we our challenge number one the mindset of the people we use AI as a tool not just simply duplicate human logic human mindset the the solution we train as engineer we have certain way of solving problem but with AI tool so the biggest challenge people trying to mimic how human solve the problem but then you did not take full advantage of AI and secondly is the trust And when AI give you a solution, how can people trust the outcome of the solution, the recommendation or decision and the the validity of such a kind of a automated decision? People still don't have the full trust on the outcome. So we still have to use a human being to verify to validate the outcome. Thirdly is a partnership. I I believe it's a challenge for every organization whether the we will see a business scenario is totally AI or is AI augmented the human engineering solution or the what is the position between human and AI this uh complex relationship I call partnership is for us to really still ongoing to discover how we can best use the human ingenuity as well as the AIN intelligence. I think that's but the biggest challenge for us is talent >> people who are uh uh we call AI native and because for people who graduate from college from the university three years ago their mindset still majority trained as a traditional science engineering approach but people who just recently graduate their mindset their approach to solve the problem are totally different from those people who graduate maybe five years ago.

[00:02:26]>> Yeah.

[00:02:26]>> So that's where we see the ability to master the AI tool, the ability to to understand the business challenge and to combine these two side. I think that's where we we see we have a need to recruit people have a both uh background.

[00:02:42]>> Professor Nate, real quick, do you have I'm assuming CL has a very dedicated AI training program for its entire workforce. Am I correct in that assumption or is it more tier?

[00:02:52]>> That's correct. That's correct. In all the employee have to go through the self-study, the company organized study.

[00:03:00]In fact, a certain department have to take a mandatory test to see the AI proficiency.

[00:03:07]>> Right. Uh Phipe, I think you have a similar program over at Schneider Electric as well, right? So, your 140,000 employees, they are also mandated to take AI training programs.

[00:03:19]>> Yeah. Yeah. From day one of this AI transformation five years ago, one of our key point was if we want to be successful, we need our entire workforce to understand enough AI. So we put a very strong emphasis on training. And just to give you one example, like many many large firms, we have this mandatory training that every employee has to do and pass every year. you know HR control typically historically we are we are training on sexual harassment anti-corruption etc we have added in the same set of totally mandatory training for the 140,000 employee of Schneider including production line workers a mandatory AI training so that they understand what AI is that they don't imagine what AI is that they don't think that AI is Robocop or that AI but they understand what it does what he not what are the limitation coming back to what professor was saying you don't trust something you don't understand so for the people to trust it they need to understand it so our trainings are a lot focused on making people understand what it does how it works and of course we do that for the 140 workers but we have also program for two top 200 people because change start from the top we have programs for people in charge of our product offerings to make sure they understand what AI can do for their customers ers we have training for the people in charge of transformation. So you need to tailor your training also for the key people and the key groups in your company that will be driving this change. But you need as I was first saying to make sure that everybody has a core understanding of what it is, what it is not, how it works so that you can build trust and what I call sometime an adult relationship with AI. Something between fascination and rejection. A normal reasonable adult relationship.

[00:05:13]>> Yeah. so important from the top to the bottom. Everyone needs to be AI proficient in this age that we are entering. Kathy, let me come to you because we've been talking about minds.

[00:05:24]This is a new AI platform introduced by the World Economic Forum recently. Uh help us understand what is the purpose of MIS. What does it stand for and who is it for?

[00:05:36]So ME stands for meaningful intelligent novel uh deployable solutions is the what economic forums global platform focused on one thing AI that actually works uh in the real world. So we're not talking about pilots not prototypes but solutions that are deployed at scale and delivering measurable results. Um, and we're trying to solve the problem of everyone's talking about AI transformation, but unless you make sure that the cases the the the kind of deployment that we're looking at are comparable, it's going to be very hard to to actually understand the impact.

[00:06:14]So, the program brings together companies, governments, and leaders who are moving beyond experimentation and want to turn uh AI into real impact. And the goal is simple to highlight what's working and share those lessons and help others uh scale faster. So we're very lucky to partner with uh you know partners such as Schneider and CL and many others uh to uh really contribute to to to this program uh to make sure the world actually has one of the largest kind of a you can call it database or repository of living brea breathing cases of uh true AI transformation.

[00:06:54]>> Yeah. And Kathy, why do you think it's important for the forum to start this kind of AI platform right now? Why is minds you think necessary right now?

[00:07:03]>> Yeah, because simply as I mentioned earlier as well, there's just a lot of hype around AI since particularly Gen AI came to the market. But what the leaders really need is a trusted platform to get inspired by applications that are actually delivering results in in practice and minds focus exactly on that. So it brings forward real world deployment and creates a space for the community also to learn, exchange ideas and openly uh discuss challenges. The goal is to help organizations skill the impact of AI through shared learning and and collaboration and we think you know only the the forums through our impartial platform can achieve just that.

[00:07:46]>> Yeah. Uh, Phipe, Professor N, let me come to you because each of your organizations of course have been recognized by the forum's minds uh, program and what's interesting is that both Schneider and CL I believe have two actual transformation stories that are profiled by the Minds program uh, whereas other companies may have one. So I guess that is a testament to how AI is changing uh, your organizations by I guess changing the solutions you provide for your customers. Professor N, let me come to you and ask about your case in particular. Give us your take on how this AI transformation actually took place. So, one of the examples is you now have an AI platform that can help you design EV battery cells. Walk us through that.

[00:08:34]>> Yes. Yeah. The EV battery is such a complex product involve multiple uh uh scientific disciplines. Not only you require knowledge in physics, chemistry, electrochemistry, thermal science, fluid mechanics and electro engineering. All these discipline interact to make a product which give you the power density, energy density and the cycle life and beside this is a a safety critical system because most of the EV battery application you sit under the passenger vehicle and it's a tremendous important to make sure the product is safe and can uh last a long life and In this case, typically a college graduate with a master degree or even PhD degree, they need to learn on the job training for three to five years before they can independently design a good high quality battery. Now with EI with the AI, we believe we can quickly accelerate that. So we actually taking all the data the the base we have created over the last 20 years or so as well as the scientific the first principle in electrochemistry in physics in the mechanical in material science we try to explore all the design space because it's like hundreds different variables and beside this product will be used in very different application scenarios working environment is also very uh uh uh varies from application to application. So in order to facilitate that design we we employ the AI tool to help us to search the feasible solutions and quickly compare the viable design and greatly shorten that. So instead of take three to five years now we can do it in minutes >> existing material system. Yes.

[00:10:41]>> So the battery design process, Professor N, before I think you were you you needed um a few weeks, maybe two weeks to do it, but now it can take 10 minutes, right? So we're not talking about a 10x uh performance. We're talking about a 2,000x maybe performance in terms of what you can do with AI. You also need, I think, to be very precise with your AI as well because we know, I guess, for LLMs there can be problems when it comes to hallucination. you might need to double check as well. So for CL's purposes, especially for something so important as electric vehicle batteries where you need safety as the top priority, you really need AI to be precise. You you can't have defects that are at the six sigma level like you mentioned. You need to have really perfection at the billion parts per product level as well. So talk to us about how important is precision AI for CL?

[00:11:38]>> Yes. uh uh give you another example in the electro coating because for a battery the EV battery or energy storage battery uh it's made of anode and cathode electrodes the coating of electrode is very precise uh let's take one example we have a anode the coating we have a the foil let's say copper foil at the five micron thickness 1.5 mters wide traveling at 100 mters per minutes and we want to code the 150 micron the material anode material and we have to control to plus minus one micron sickness variation and traditionally rely on human being no matter how hard how diligent the workers it's very tedious very demanding work they cannot control to a very tight tolerance so with the AI tool We allowed the system to auto autonomously adapt to the process of fluctuation because there will be batch to batch version. There's a viscosity of the slurry will change depend on temperature. There are so many uh factor could affect the coating uniformity. Now with AI tool we can do much better with two orders of magnitude better than best uh uh operator can do. So that is one example how AI can help. Now you mentioned about the six sigma. In fact for our industry six sigma is not sufficient because for most of other industry for semiconductor electronics automotive applications six sigma is the best inclass in terms manufacturing quality. Uh since our product is a safety critical system and each vehicle in order to guarantee vehicle level ppm part per million uh defect rate we have to demand our battery to be at the ppb level per billion uh uh tolerance. uh in order to achieve that really there's no human being can manually assure that kind of uh precision and repeatability and uniformity and therefore we adopt a large uh amount of uh AI solutions from the quality uh uh uh the root cost analysis to the process control to all these uh process parameter optimization.

[00:14:16]So all these uh are using the AI that is sometime I call EV battery manufacturing require tender loving care because so delicate and yet subtle variation could give you a huge problem down the road.

[00:14:34]>> Yeah, I guess that is the power of AI right now. There are so many thousands I guess of interdependent factors when it comes to producing an EV battery and your AI has to be powerful enough to make sure all of those check out where if one variable does change it doesn't affect the other variables. Philipe let me come to you and let me ask you about Schneider's uh AI transformation power because right now you have two solutions being profiled by the mines program. One is about using AI to optimize the entire building's energy management system, but also using AI to optimize the energy management system of an individual room as well. Talk to us about these two individual use cases and how AI is transforming the energy management system.

[00:15:16]>> Yeah, with pleasure. So, first I would say that minds is really a reflection of what we decided to start five years ago.

[00:15:23]So in Schneider we had since the 80s of course like any large corporation machine learning teams working on AI and on technology. Five years ago we said it is time now to really accelerate and impact at scale our customers to do that at scale as I was saying earlier we start from the business value. So our customer needs did not change a lot since 10 years. They want a cheaper energy. They want to use less energy.

[00:15:50]They want to use a more decarbonated energy and they want to do that in an easy and simply and simply and and simple way and I will give to illustrate my example. So our transformation was to build process and teams and organization and maybe we'll come later on that to deliver that. But if I illustrate that by a very concrete example something that looks as simple I would almost dare to say as stupid as a temperature room controller in a room. The value for the customer is simple. They want comfort, the right temperature, and they want it with as little energy spent as possible.

[00:16:29]10 years ago, we have connected these room controllers for office meeting rooms, for hospital rooms, for hotel rooms, to the building management systems to heat and cool only when it's needed to save energy. But then what we did four years ago was going to our customer and they told us we want even more comfort. We want even more energy savings. So the next step was to say let's say you want your studio Michael to be warm at 10:00 a.m. in the morning or cold if you're in summertime. The next question is when do you need to start your heating or your cooling? 10 minutes before 10, 5 minutes before 10, 30 minutes before 10. We embedded an AI algorithm in each of our room controllers. So we talks of millions of PCs all over the world that learns the behavior of the room, the thermal behavior of the room and is able to optimize and decide exactly when to stop, when to start, increasing comfort and removing between 20 to 30% of energy cost. So that sounds simple. This is not a very large genai cloud solution, but that's a solution that truly impact our customers wallet. They can spend 20% less cost on the energy of heating and cooling buildings. It impacts positively the planet because they use less energy and you can deploy that at scale without anything. You need an an electrician, a screwdriver, put it in the wall and you say 20% energy. So for us that's an example of what we call AI at scale that truly impact the planet easy to deploy massive impact positive on the planet positive on the on the savings. Very simple very easy but that's what AI at scale is >> right. Uh, Phipe and Professor N, I want to ask both of you what kind of perhaps roadblocks, inertia that you might have faced, frictions within the organization to push these solutions out at the organization level because right now it sounds like these two solutions are now scalable. They're fully deployed. But was there organizational inertia friction in terms of coming up with this business challenge and solution that you really needed to convince stakeholders within the organization? and how did you deal uh with really pushing through this idea? Phipe, let me let me start with you and then I'll turn to professor non the same question.

[00:18:48]>> Absolutely. Back to what professor was saying, one of the obvious difficulty is to deliver a solution with AI at scale that impacts you need to find a way first to merge the domain knowledge, the understanding of your customers and the AI knowledge. And the reality is it's almost impossible to have the two in the same person. People with high knowledge are out of university since let's say five to 10 years max when to build knowledge in electrical complex system you need 20 30 years of experience. So the first thing we had to solve was find a way to merge domain knowledge with technical knowledge. So we created a central team with a very strong AI knowledge bit more than 400 people data scientists data engineers etc. Then we work with our businesses to identify what are our customer needs, what are our business needs, where is their value for the planet, where is their value for Schneider and for our customers. We start from this business value and we create a team following the principle of agile where we merge AI specialists coming from my team with domain specialists coming from the domain team.

[00:19:53]So that's the first thing. The second thing is we understood when we started that many many companies many many people are stuck at the pilot phase. So we made the bold decision to say this team will not be in charge of developing a pilot a proof of concept but that team will be in charge to deliver a fully scaled solution in production. Which mean that when we do a pilot we do much more than checking the technical feibility the data availability. We check the business case. We check the customer acceptance. So I could speak forever but if I make it very very short we have have tried to take AI from innovation into a kind of almost normal R&D work where we have a team which is in charge not only of proving it can work but making it work at scale from ideation to production and to do that we merge domain knowledge with knowledge into an agile team and that was the best way to solve what people often call the pox syndrome the pit full of demonstrators where you only prove it's feasible. By doing that, we are at scale, we're in production, we're delivering. Not super easy to put that in motion, but we've been starting five years ago and it's working pretty well.

[00:21:06]>> Yeah, Professor N, would you agree with what Philipe just said? And was there any kind of organizational inertia on your end in terms of pushing through the AI uh design and AI manufacturing platforms that are currently now a CL?

[00:21:20]Uh yes we in addition to what Philips mentioned at CO we see another inertia is in a mindset. Uh actually CAT is still in the rapid growth uh uh phase and employees are not worried about their job being replaced by air but people are thinking ways to improve their productivity because the the business expanding. I think that the big inertia is a mindset as I mentioned earlier and people try to solve the problem even with a new tool but they still want to solve the problem in the old traditional way. I think this is where emerging that take time to train people to get used to this new way of problem solving. uh to to support this CAT actually established not only at the headquarter in India but the team uh research institute in Hong Kong research institute in Shanghai try to build up this kind of uh in-house expertise to uh develop a more user friendly tools for our engineer to use >> right uh Kathy let me come back to you because we just heard two fantastic use cases of AI and how AI is transforming business solution solutions at Schneider and CL one of course is a European company based in France another is a Chinese company Kathy how is the regional momentum in mind right now particularly regarding China as well >> yeah the regional momentum in minds is strong overall but greater China stands out very clearly uh just to give you some numbers nearly half of all global finalists and around 50% of uh short list uh shortlisted applications s come from greater China which reflects not just scale but also the quality and maturity of the the deployment. In the latest cohort alone, about 23% of applications include at least one organization from the region with strong participation from both large enterprises and uh innovators. And what differentiates China at this stage is the ability to move from innovation into real largecale deployment. Uh similarly along the lines of what uh Philillip and and and professor N was just saying, we are seeing uh AI implemented across major industrial and uh enterprise settings including energy manufacturing and finance with companies like uh CL uh state grade um ICBC um and and Foxcon along uh alongside other emerging players. Um this momentum is supported by a combination of factors. Uh clear policy direction, strong industrial ecosystems, advanced digital uh infrastructure, access to data and talent and a large uh receptive market.

[00:24:17]And together this created the conditions not just to experiment with AI but to scale it uh quickly and effectively. So the key takeaway is that China's momentum is less about isolated innovation and more about execution at scale and this is what um this is what is um uh making it one of the most important regions to watch in the next phase of AI deployment.

[00:24:41]>> Yeah. So professor N like Kathy was mentioning there's a lot of representation in the minds program from companies here in greater China. What do you think is driving China's momentum in terms of moving from innovation to largecale deployment more quickly?

[00:24:59]>> I think China is one of the few countries in the world has a full scale of industries. There's lot of uh production site and the data is the unique competitive strength. So other result many of the business large or small and they they know the AI tsunami is coming and everyone is trying to find a way to uh be to ride the wave inside be left out by the tsunami. So other result people are trying also the local government encourage business to try and deploy new way. In fact, early this year, uh I think eight the department in the central government issued a common uh guideline to encourage uh AF plus and industry. I think that's in line with Cassie's World Economic Forums minds program that the government actually issued the guideline to encourage all industry to embrace this new uh uh technology and try to see improve their uh business competitiveness.

[00:26:14]>> Yeah. And Professor N just a quick followup. How would you evaluate I guess the intersection of academia and industry here in China? Is it unique in any way? uh how does it contribute to China's overall innovation ecosystem?

[00:26:28]I think if you look at yes, China uh just like United States has a large number of universities and institutions and traditionally I think US is strong in the basic fundamental research and Chinese university on the other hand is doing more are stronger in bridging the gap between fundamental research and the industrial application. I think that scale also China produce more STEM graduates. Uh in fact probably China every year China produce 13 14 million graduates. More than half of those graduates are STEM majors. Uh on the other hand the US is a much smaller maybe 10% of that uh number. uh that gave China's business uh some uh really uh I would say uh grassroot uh unique strengths for them to embrace this technological revolution >> right uh Philipe let me come to you because Schneider Electric you operate across more than 100 countries and regions and I think that really gives Schneider a very unique comparative view the Chinese market has been compared to the world's toughest gym for businesses because innovation here in this country happens at such a rapid pace. What can companies do you think in Europe maybe take away from the speed of AI deployment that we're seeing in China's industrial sector? And conversely, Phipe, what do you think Chinese companies can still learn from global best practices?

[00:28:05]>> Yeah, so we in Schneider, you know, we we are not anymore. I mean we are very global company multilocal and we really take pride in our system where we are again as local as possible everywhere. China for us is big. Yeah.

[00:28:20]China just to give you a rough idea it's 18,000 employees it's AI laboratories it's five R&D centers 23 factories many many customers. So for us China is a very important region and in that case AI plays a very important role also. uh and I would say that if I see one place where China differentiates is probably a lot on the speed of adoption in our factories and in the solation sector uh where we see a very very fast deployment of AI at scale uh in our factories here in China and that we leverage globally uh that's certainly one one key things then for me you know I'm a strong advocate of all of our regions all our AI teams collaborating together for brighter future So we try to avoid to try to have a too much local view etc. but really bring knowledge habits common things together to develop AI faster and for true impact at scale. So again for us China one of the key pillars of our strategy uh with Europe with US with India fundamental for success and super important. Yeah, but Philipe, are you seeing maybe perhaps the innovations of uh the Chinese ecosystem? Can you use that for your global operations as well? So for your R&D centers here in China, can you apply that for your global operations?

[00:29:42]>> Definitively, we leverage what we learn in China across the world. Uh as we leverage in China, things that we learn in other places also. But yes, definitively there is a very strong cross fertilization uh of of AI between those different regions. Absolutely.

[00:29:58]Kathy, let me come back to you and let me ask you more about minds because there are many AI initiatives globally.

[00:30:04]What do you think makes the minds program stand out?

[00:30:08]>> Yeah, I think first and foremost uh we will need to go back to the value proposition of the word forum which is truly multistakeholder. I don't think there's another platform that can gather this many multistakeholders like the forum does. So that's first and foremost why it's differentiated and the minds itself is built on a rigorous review process that ensures the cases that we spotlight are tangible proofs of uh AI delivering uh real world value. So what that means in practice is we bring together leading experts and thought leaders leveraging the platform that we have across the world to assess and validate this uh transformations and beyond selection. We also created a community where organizations come together to share learnings and experiences not just what works but also what didn't work because that data is also extremely uh precious. So it's not just about identifying strong cases, it's about turning them into a collective insight for scaling AI impact.

[00:31:10]>> So Kathy, now there are what two cohorts so far for minds. Uh what do you think are the key learnings from these first two cohorts?

[00:31:20]>> Yeah. So we go beyond showcasing. We extract what works and shared it across the community. And some of the key lessons are I think already earlier you know professor N and um and uh Philip already touched on some of them. So first of all we need to treat AI as a core capability of your organization not simply as tools. uh we need to uh invest in both technology and people because while technology is critical, it only creates value uh when organizations have the right skills, processes and ways of working in place. Um and then last but not least, we need to build trust early especially in highstake uh environments.

[00:32:04]Again, just very much demonstrated by the cases that Professor N and Phillip uh gave earlier.

[00:32:10]>> Yeah. Uh, Professor N, let me come to you and ask you about how important is the value of data for CL. I want to ask Phipe this question as well. It might seem like really the obvious part in terms of how we want to use AI to really transform business solutions, but you really do need quality data. And one of the key findings from the forum's report released earlier this year called proof over promise is that data perhaps could be one of the biggest barriers to having good AI solutions for businesses.

[00:32:44]Professor N, how are you thinking about data quality at CL?

[00:32:49]>> Uh yes, definitely the data quality is essential to the successful deployment of AI applications.

[00:32:58]uh however uh currently the AI models the larger models already has enough uh uh knowledge acquired from the open data open sources uh I think every individual business if you try to apply to a particular sector and then the domain specific data becomes critical because the AI the gen AI it's like a college graduates but if you want that AI to be able to solve your toughest uh uh engineering or industrial problem then you need at least master PhD degree level intelligence then the additional data from the company uh become valuable to to fine-tune to optimize the model so that the model can generate some relevant uh solutions. uh again this model uh the data is critical but I want to emphasize the usable data many company face a challenge they have lots of data but if you look at the data they mostly in the isolated I will call data links and how do you merge all the data into a large ocean that is a challenge >> yeah and professor N you're also developing what's called physicsinformed AI right so it can't just be data that you have from the computer for your purposes for CL is that the next level when it comes to embedding AI into the organization to have physics informed AI maybe explain to our audience what that means >> uh yes that's true our our founder chairman and CEO wanted to develop even for the next phase AI we want to develop a AI uh Einstein and be able to really to make sure to master all the science, maths, chemistry in the system. So we want not only just look at for our own internal use but we want to develop more general purpose uh uh AI Einstein >> right an AI Einstein. Uh Philipe I'm wondering if you have a similar ambition over at Schneider as well. So you are the chief AI officer but you also have a chief data scientist and chief data officer at Schneider as well. So how are you thinking about data quality when it comes to providing business solutions for your customers?

[00:35:22]>> Yeah, definitively the data and the specific domain data and this understanding of electrical data etc are critical for the success of AI. We cannot just rely on general models or we just provide general solutions. Uh and the effort on data is has started has to be continued. But I would like to highlight one thing is all our companies we spend a lot of time improving the quality of what we call in the industry structured data databases tables etc. With generative AI you have no access to what is qualified as unstructured data text documents images all of that. So one of the first challenge we are all facing now is how can we bring the same rigor the same quality on the management of unstructured data than structured data.

[00:36:12]So that's one big axis of effort. The second thing I would like to say is of course it's important to clean and organize your data but it's an endless fight. So sometime I say first choose what is the use case you want to solve.

[00:36:27]What is your business need? What is the AI that you need? what are the data that this AI need and then focus on organizing and tuning those datas.

[00:36:36]First, it will narrow the scope. Second, it will bring a closer ROI to the effort on data. And third, it gives meaning for the employees to whom we are asking to clean that data because they know why they do it. Cleaning for cleaning can be a bit difficult. Cleaning for a given outcome brings value, brings speed, makes things much more efficient. So yes, the work on data is critical, but let's do work on data with a clear purpose in mind for what case what AI application instead of trying to solve everything which can take ages and ages and delay a real world application of AI. If you start want to clean all your data, you're going to start AI in 20 years and you'll be late already. You see what I mean? So that's what I would say about the importance of data, but definitely super clean and super important. uh and when I hear what professor N was saying on on Einstein on our side the challenge is is is similar but different of course what we want is to give to our customers what we call energy intelligence so we want to give them a complete understanding of what's happening in the energy systems give them the ability to say hey you know what the biggest priority for me in the next months is saving on cost or the highest priority is reliability or everything make me the best recommendation optimize from the set points. That's what we want to give really. We want to give our customers ability to fully understand and act as autonomously as possible with AI on optimizing the energy system, what we call energy intelligence.

[00:38:09]>> Again, start with the business value.

[00:38:10]Again, start with the business case first and then think about the technology. Uh Kathy, let me come to you and give you the last word because one of the best things about a program like this is we can preview the annual meeting of the new champions of the World Economic Forum which will be held in the beautiful seaside city of Dalian and northeast China this June. Uh Kathy, share with us the overall theme and the key AI related priorities that the community should pay attention to during this year's annual meeting of the new champions.

[00:38:39]>> Absolutely. We're very happy to be back to Dallian. Um the theme of the annual meeting of uh new champions in uh 2026 is innovation at scale. Um the pro the focus is on how innovation especially AI translates into real economic progress.

[00:38:58]Um this is less about what AI can do but it's more about how it's already reshaping industries driving enterprise transformation and creating uh growth.

[00:39:07]From a AI perspective there are three priorities to watch. First, deployment at scale. The conversation has already moved beyond uh pilots. The key question now is how organizations turn uh AI into repeatable value across core uh functions. Second, enterprise transformation. The biggest gains are coming from redesigning workflows and operating models and not just adding uh AI on top of existing processes. And third, the foundations that make scale are possible that includes infrastructure, data, energy, governance and talent. Uh these are now real constraints but also the biggest uh um enablers. We'll also see the next phase of minds which continues to highlight what is um actually working in practice uh across regions and industries. So stay tuned for the uh for the third cohort uh announcement that's going to come out soon.

[00:40:03]>> Yeah. And I would be remiss Kathy not to finally ask you how organizations do you think can better engage with minds and the broader AI initiatives that we see at the forum.

[00:40:14]>> Yeah, very simply uh if you're a partner, non-partner, please do uh apply directly through the M's website. The uh process is quite straightforward. Uh you can submit a application that shows how how your AI solution is already delivering real impact. From there it goes through a review and validation process with independent experts. The strongest cases are selected based on real results and their ability to scale.

[00:40:40]It's open to both uh public and private uh organizations that are already deploying AI and can demonstrate measurable uh outcomes. So not only private sector but also government uh as well.

[00:40:54]>> Okay. And we are going to leave it there. That is all the time that we have today. I think if there is one takeaway from this conversation, it's that AI is no longer just an IT project. It is of course a test of leadership. It's a test of organization as well. The technology is available to everyone, but the competitive advantage I think belongs to those who are really willing to do the hard work of redesigning their processes, empowering their people and also building trust by design into their AI ecosystem. So, a massive thank you to Kathy Lee, to Professor Nijun, and to Philip Hbach as well for showing us what it truly means to move from promise to proof when it comes to AI. And thank you all for joining us. We'll see you again next time.