AI needs Energy and Energy needs AI

00:00:00: A very warm welcome to Peter Weckesser, Chief Digital Officer at Schneider Electric.

00:00:05: Peter please introduce yourself to our audience.

00:00:09: Yeah first of all good morning.

00:00:11: thanks for having me today.

00:00:13: my name is Peter Weckscher.

00:00:15: I'm the chief digital officer at Schneide Electric and we will certainly talk a little bit more about what i do in my role as chief digital.

00:00:29: Title states, AI needs energy and Energy Needs AI.

00:00:33: That's a fascinating combination!

00:00:36: What does that phrase actually mean in practical terms for business and society over the next decade?

00:00:45: Yeah let me maybe start with the first part of that statement – AI Needs Energy which is more obvious it very much in public debate.

00:00:57: The build-out of our AI infrastructure, the build out of data centers and AI factories is actually constrained not so much by the number of GPUs available but it's very much constrained by access to energy.

00:01:14: Energy became a single biggest challenge in this technology that will influence most likely every industry and every company pretty significantly.

00:01:27: So it's a well-known fact that AI is energy hungry, AI needs a lot of compute And we will certainly go into how efficiencies are being achieved But its simply a fact It needs a lots of energy.

00:01:45: in this energy need to be provisioned.

00:01:48: Now what does take to provision the amount of energy?

00:01:51: First off all these energy needed to generated.

00:01:54: And there's of course all forms of energy being used today, fossil-energy generation renewables nuclear.

00:02:02: Everything is in the mix and it's very clear that there's a requirement from the AI side, more energy.

00:02:12: and of course best energy is sustainable energy from renewable.

00:02:19: That this gets created at scale.

00:02:22: second that energy needs to be distributed, and also there are very significant new requirements towards the distribution grid.

00:02:32: And this is very much connected to build-out of renewable energy resources which mainly wind and solar.

00:02:40: Wind and solar are highly distributed.

00:02:43: they come with pretty significant new requirement for the grid that becomes in future way more complex.

00:02:51: it has become intelligent.

00:02:54: Now this takes me a little bit to the other side of that statement, that energy needs AI.

00:03:01: If you think about the grid in future and maybe I describe old grids as new grids on super high level In the past –and still is the case today– we have relatively small number very large power plants.

00:03:20: The electrical energy that's being generated first gets transmitted and then it goes into the distribution grid, media voltage low-voltage distribution grid to be made available for all consumers.

00:03:36: So this is basically a one-directional energy flow from the energy generation to the energy consumption.

00:03:43: Now, the new grid becomes way more complex and we all know that solar on rooftops large solar farms are being built out at significant scale.

00:03:58: Also, wind is being built out and these are highly distributed.

00:04:01: Much smaller amounts of power generation but highly distributed in aggregate.

00:04:07: In most geographies in the world renewables represent by far The most significant addition to our electrical energy generation.

00:04:18: Now the nature of these renewables Is that they will have a significant demand To build up the grid as well.

00:04:25: So because they become part of the distribution grid immediately and that means we need to make significant investments into our distribution grid.

00:04:38: Now, as it was built many years ago for maximum capacity... As you put more load on your distribution grid two options, or the combination of two options.

00:04:52: We can build out this distribution grid and basically put more copper into the ground.

00:04:58: This would be the most expensive way because we will again have to design that distribution grid for its maximum capacity And designing it for peak.

00:05:11: Now what you really need is make your distribution grid intelligent Because now energy flows between all these producers of energies, wind turbines solar panels and this might be large solar park so it might be a solar installation on our home.

00:05:30: And that means these energy flows need to be controlled.

00:05:34: let me bring one more aspect into this.

00:05:38: renewables by nature, do not provide energy at any given time of the day.

00:05:45: Of course you need sunshine and wind for renewables so you have to store energy as well.

00:05:52: You also build storage capacities.

00:05:54: This is battery parks or electric fleet, automotive fleet that are growing in order to store their energy.

00:06:05: From this description we can already see that actually every home in the future becomes what we call a prosumer.

00:06:13: A prosumer is an artificial word of consumer and producer, which means every industrial site will be generating energy but also consuming energy.

00:06:26: so at certain times this asset or building on the industrial side will provide energy to the grid.

00:06:38: Robbie, I take a quick pause here for you to give me your next question.

00:06:44: Peter thank you very much for the holistic introduction and big picture.

00:06:49: when we tried to assess this little bit what do you believe?

00:06:53: are biggest energy infrastructure bottlenecks emerging?

00:06:59: Furthermore how can industry address them properly?

00:07:05: Very good segue probably in the next part, but I want to go into a little deeper.

00:07:12: So The Public Debate is very much about sustainable power generation and renewables so The public debate is less about demand site management.

00:07:22: And this where second part of introductory statement energy needs AI comes into play.

00:07:29: If you think about evolution of distribution grid that i that I have been describing, then it's very obvious.

00:07:37: that all of a sudden we have to deal with the complexity of the grid which has millions of participants, millions of these prosumers being part of this new grid infrastructure.

00:07:49: That's not something for future but it is already happening today!

00:07:53: Now in order to stabilize the grid utilities will have the challenge they'll know from each participant on their distribution grid how much energy do they need that means in the next fifteen minutes, or how much energy will they be able to contribute.

00:08:15: Now you can imagine this is a non-deterministic problem where you have millions of participants and this way I will be playing a major role because there is not a deterministic control algorithm that can deal with this kind of complexity.

00:08:31: By the way, the grid structure will change on daily basis new participants being added to the grid some participant potentially been removed from the grid.

00:08:42: so in basically every participant that is your home but it's also an industrial site will be asked by the utility, by the grid operator to provide not only current consumption or current production of energy, but also providing forecasts for future consumption and production.

00:09:03: And this leads us into the demand side management and also to the management on when do we need to store energy in the battery capacity that we have available?

00:09:14: As you can imagine AI is starting to play a major role here millions of data points that provide current consumption and production, millions of data points that provide future production in consumption.

00:09:28: And from there the AI will be deriving right decision making which is should we charge battery capacity?

00:09:38: Should we drain it because they have so much energy need now not enough wind or sunshine!

00:09:47: Or should we also manage the demand side of large consumers?

00:09:52: Of energy, very large industrial sites that have some flexibility when they can consume their energy.

00:10:00: AI will be playing a major role in this and I'll give you an easy example as a consumer.

00:10:07: probably most people today or in the future would be able to relate too.

00:10:12: if you think about your home And if you have a home where maybe your solar panel is on the rooftop, of course there's a grid connection.

00:10:22: You might also have an electric vehicle in your garage!

00:10:27: If you think about your own energy consumption it's ideal that you consume all the energy generated by your solar panels directly into your house And there are a couple of scenarios that could happen.

00:10:41: Sometimes you don't produce enough energy, so you need additional energy from the grid.

00:10:48: sometimes You produce more than you have.

00:10:52: then probably two options maybe three options certain workloads that consume energy in the home.

00:11:00: That's maybe not totally realistic to do that, but you could start your washing machine and other appliances to consume more energy.

00:11:08: if you have access of energy You COULD charge your car which would be probably the smartest way To store that energy because If you put it into the battery Of your car...you can either use that energy for driving Your car or you could discharge your car at times when there is a lack energy.

00:11:27: And the last option is to feed the energy back into.

00:11:31: Now this seems to be like a pretty complex scenario, right?

00:11:34: Who makes the decision.

00:11:35: You don't want an app on your phone where you basically make these decisions online in real time but we have very smart algorithm that makes those decisions for us and of course it does not only need to know about energy production consumptions or future forecast behavior, right?

00:11:59: When do you want to use your car.

00:12:01: Of course when you want To Use Your Car You'd Prefer To Have A Charge So That You Can Go Where You Want And so Bringing This All Together Will Be Based Very Much On AI Algorithms To Manage The Microgrid In Your Home.

00:12:16: But Then AI will Also Play a Major Role in the Interaction of Millions Of Microgrid Within The Distribution Grid.

00:12:24: Yeah, thanks for this explanation Peter.

00:12:27: it shows really the necessity for AI because I think nobody will Will like to manage that manually.

00:12:36: That leads to the question.

00:12:38: I would even say possible to manage it manually right?

00:12:40: so the manual overrides?

00:12:42: I think a manual override will be possible But I think it will be only used in exception cases.

00:12:49: Peter historically technology becomes more efficient by time but overall usage increases.

00:12:58: So do you expect AI to reduce energy consumption?

00:13:02: Overall,

00:13:03: or

00:13:04: will the efficiency gains be outweighed by the explosive demand of computing?

00:13:09: what seems

00:13:12: So?

00:13:12: this is a very interesting question, which has of course widely debated.

00:13:16: All the forecasts that I'm aware off and also our own forecast at Schneider show in terms of energy will be significantly growing.

00:13:29: It's actually the fastest-growing industry, In terms of Energy Demand So that energy demand is predicted to double within the next couple years.

00:13:39: Still today The total consumption Of energy out of global consumption for data centers Is only at four percent But it's growing rapidly.

00:13:50: Now This is all of our demand for AI, so everything that we do with AI.

00:13:56: It's it's... All the AI that be used for software engineering or For all the cool large language model tools That you use on a daily basis.

00:14:07: You have perceived that internet search Is becoming very much AI based and supported.

00:14:13: So this consumption of AI drives a higher energy demand for AI.

00:14:20: So this goes without saying, but it's really across everything that we all do and in the cross-everything that you'll use AI for Now.

00:14:31: at the same time technologies off course improving so you could look at GPU As an energy to token converter, right?

00:14:42: So that's a very simplistic model.

00:14:44: how too.

00:14:45: How do you get it at the GPU?

00:14:47: and of course this gets better over time.

00:14:49: so GPUs become more efficient per unit of energy that they consume.

00:14:58: And every generation will become better, and by the way it's not only the GPUs but also the whole data center gets better in its energy supply infrastructure or cooling infrastructure.

00:15:11: next generations of data centers fully go to liquid based cooling which drives a higher level efficiency.

00:15:19: so it is totally clear technology get better over time.

00:15:24: But it's not only the hardware technology, so its' not just energy supply and cooling or GPUs itself.

00:15:30: It is also software algorithms!

00:15:32: All of these companies are part of that AI stack which starts with large language models but then algorithms on top become better in terms how many tokens they need to consume.

00:15:50: I see that the industry is driving significant energies and that ratio of token per watt will get better in the future.

00:16:01: But it's also very clear, The amount of tokens we all consume for all AI based applications continues to grow pretty significantly.

00:16:11: What we talked earlier about, that you also will be using AI for the optimization of our overall energy infrastructure.

00:16:21: This is very clearly a positive equation – The amount of energy it takes to generate tokens and then optimize the grid is a very good investment in sustainability.

00:16:36: Thank You Peter!

00:16:38: All of your answers already provided a solid foundation for our audience, but let's try to make this a little bit more tangible as well.

00:16:46: So where is AI already creating measurable value in energy systems today?

00:16:51: Can you share a few examples?

00:16:53: not in pilot projects... ...but in real-world

00:16:56: operations?".

00:16:58: We have a good number of real world operations that we use AI to manage microgrids.

00:17:05: so There's an example of a very large shopping center mall in Finland that has deployed full microgrid installation with on-site power generation, battery capacities and demand site management.

00:17:29: It was decided to use AI based algorithms from the very beginning To drive optimization of Of the whole installation.

00:17:39: So that's tangible value.

00:17:41: Then you can see, I want to give two more examples.

00:17:46: This is coming in the home space and we talk about the Home Energy Management Systems or the HEMS controller which will all seen our homes in future.

00:18:00: this an intelligent device that it'll go into your electrical cabinet And this is not totally, it's evolving right?

00:18:09: But the technology is available.

00:18:13: You can buy these from us and others.

00:18:21: of your home when it comes to the consumption and distribution of energy.

00:18:27: This HEMS controller does exactly what I was explaining earlier, that controls the energy flows between your solar installation –the grid– the battery in your electric car optimizes the energy consumption for your home which results into real savings!

00:18:44: When i say this is available at homes available for buildings, commercial buildings and office buildings.

00:18:52: And also multi-apartment buildings.

00:18:55: this is already available today as well.

00:18:59: Last example It's actually data centers.

00:19:02: Data centers is super interesting, right?

00:19:03: Data centers are very active participant to the elective distribution grid.

00:19:08: They are power hungry they have a large consumption of power And we're one of leading providers of electrical and distribution and cooling infrastructure into these data centers.

00:19:23: We work with all our customers.

00:19:25: optimize this energy consumption in the data center.

00:19:29: And by the way, this is actually the same technology that we also use for predictive maintenance of all the assets that are in the datacenters.

00:19:37: This is our equipment That we sell to the datascenter operators and serving two purposes really optimizing The energy flows in the Datacenter depending on the load that's under on the datascent and also providing predictive maintenance information on which asset needs maintenance at what time because data center operators certainly have this requirement of a hundred percent availability off their assets.

00:20:05: Peter you spoke about homes, datacenters and office buildings.

00:20:11: how bout industrial companies who are producing?

00:20:15: where is the biggest untapped opportunity at the intersection of AI and energy management for this industries?

00:20:24: So look, industrial companies are a bit similar to data centers.

00:20:29: The data center is now called AI factories right so it's a factory that produces tokens.

00:20:36: um And if you look at other factories no matter if this is automotive or chemical they all need energy a wide scale of very, very different products.

00:20:53: Now all our customers also on the industrial side have that challenge to optimize their energy consumption which is Using the right energy mix, and that's a mix between on-site generation.

00:21:10: That might be with combined cycle plants or solar or even wind.

00:21:15: but all of our industrial customers also still have a grid connection today.

00:21:19: So an industrial site is microgrid which needs to be very actively managed.

00:21:25: All principles we debated for homes and data centers before apply exactly in the same way for industrial production sites.

00:21:32: And industrial sites certainly, depending on the nature of your processes have the capability for demand site management.

00:21:42: So certain industrial processes that particularly ones that consume very high amounts If the customer can, you start these processes at a time where there is an access of energy available in the grid because this energy becomes much cheaper.

00:22:01: So if you have a windy day with sunshine that's ideal scenario to start high energy intensive processes, like take an aluminium smelter for example which you can operate of course only with large amounts of energy.

00:22:18: and then this energy.

00:22:19: You want this energy to be as cheap as possible.

00:22:23: This is happening at times where there's an axis of energy.

00:22:29: Many of our industrial customers, this part of the demand side management that you start your energy intensive processes at the right time becomes a key topic.

00:22:39: And let me just make one add on to this actually some of their energy-intensive but very slow process fall into that category as well.

00:22:48: This is particularly heating and cooling processes for buildings.

00:22:54: A heating process is usually a very small process.

00:22:57: You can start that an hour earlier or an hour later, and the building itself... will preserve that energy at least for a certain amount of time.

00:23:13: And using this, for the very slow processes you start them and stop it in the right time is an important part of their demand-side management.

00:23:21: This would be controlled by AI because you do not only need your production schedules but also to have forecast on when are the energy prices being lowered so we can calculate really optimum times to start our business process.

00:23:35: Peter, it's tried to connect a few more dots.

00:23:39: Before AI became such an important topic we talked very much about digital twins and simulation models over the last decade or even for longer period of time.

00:23:51: You mentioned predictive maintenance and the similarity of tools.

00:23:54: already Schneider Electric has invested heavily in digital twins and industrial software And how does generative AI support the role of digital twins in managing and predicting energy-intensive operations?

00:24:12: Very good topic.

00:24:15: Let me go into this a little bit.

00:24:17: so far, we have mostly talked about the Operations phase.

00:24:21: now if you think of Digital Twin We are starting to talk about The Engineering Phase.

00:24:25: So at Schneider our own capabilities for the design of the electrical system, design of an electrical grid and also very strong simulation capability.

00:24:41: We have this same with our software company Aviva that has strong capabilities on the process engineering side.

00:24:50: So if you think of this, AI is starting to play a major role in the engineering process.

00:24:56: While in the past let's say designers have an electrical system started to design with graphical desktop tools it was called single line diagrams which basically are the representation of digital twin of an electrical systems Now we see more and that AI tools become a very, very relevant part of the engineering process.

00:25:25: To really have much faster time from idea to fully build and fully simulate a digital twin.

00:25:35: Let me go into this simulation part because it's key already something we are offering many customers today.

00:25:42: Well then the engineering processes is usually concluded by a validation and simulation that tells you, hey yes my digital twin does exactly what it's intended to do.

00:25:55: And once that's validated we can start building the real asset.

00:25:59: Let us say this asset is full electrical distribution system of a data center.

00:26:08: Now, as we then also do not only provide the electrical design for the data center but we provide a lot of equipment.

00:26:16: that's low voltage distribution assets.

00:26:19: The cooling assets.

00:26:20: this is all part of the initial digital twin and Then we help our customers to build the real asset.

00:26:27: That of course puts us now.

00:26:28: are customers in into that position that you can always sudden Not only take the real data from the real assets and try to build again physical models, but we can use simulations that have used to simulate a digital twin also for the real asset.

00:26:46: So this puts us into position where you really bring together the digital twin in the real twin and compare the simulation results of the digital twins with the real life result.

00:27:00: It also puts us into a capability and we call that energy intelligence to interact with the real digital twin in very different way.

00:27:11: And under this, I want give you maybe one or two examples on what do mean by energy intelligence?

00:27:17: We can now ask natural language questions through The Real Asset asking when is next maintenance required and what is the best timing for that next maintenance.

00:27:30: And give me all the assets, that need maintenance in the next ten days right?

00:27:34: This was not possible in the past.

00:27:36: now we can ask these kinds of with a help over AI based on top of our operations platforms.

00:27:43: We can asked this natural language questions to get it full list.

00:27:47: Hey!

00:27:47: These are the fifteen assets That needs maintenance In the next fifteen days.

00:27:54: But we can also ask very different questions.

00:27:58: What if you have a failure in one subsystem of the electrical distribution, and by the way electric distribution within data centers is always redundant?

00:28:10: Now you go through scenarios where say what if he had a failure on this medium-voltage subsystem?

00:28:19: Can we still sustain the operations of the data center or do we need to have emergency power supplies to kick in?

00:28:27: So, you can go through all these kinds of simulations while the datacenter is running and of course the AI is now capable.

00:28:34: In case if you really an issue that a subsystem has a problem that within fractions of seconds, you can compute the scenarios in what if model and then take most appropriate counteractions.

00:28:53: So AI plays major role in engineering process building their digital twin but connecting it to real twins using design data together with operations for AI.

00:29:09: So Peter, now that we have built the Solid Foundation.

00:29:13: We talked about pilot projects and real world operations on top of them in the industrial space also domestic buildings.

00:29:27: How close are we to a future where buildings, factories and microgrids autonomously optimize their own energy consumption using AI?

00:29:35: And by how close?

00:29:37: I mean more than eighty percent are using it seamlessly.

00:29:41: So very good question and not easy to predict right?

00:29:46: so let me give you the answer in two dimensions.

00:29:50: The technology is moving at rapid speed.

00:29:54: many of these asks are actually doable with today's technology already, right?

00:30:01: So if you use the latest technology that our hands already today we can build autonomous microgrids.

00:30:09: That self-optimize The energy consumption of a defined asset.

00:30:15: now adoption is probably the bigger challenge because this requires investments.

00:30:21: This requires clear return on investment calculations as well And some of these industries are moving relatively slow.

00:30:31: If you look at this from a global perspective, we can see that China is moving faster than anybody else into the direction and other geographies are much slower then China for example.

00:30:45: In Europe... We're also slow much slower than China and I think we need to push on the adoption of these technologies because they are clear intangible benefits already available today.

00:30:58: Yes, there's some investment required.

00:31:01: just to give you one number... We're talking big numbers.

00:31:06: if you think about the build-out of the distribution grid in Europe that is only EU Europe Then the forecasted number is that investment required will be seventy billion per year over the foreseeable future.

00:31:24: So pretty significant investments, but it's an investment where Europe cannot stop or even wait, because out of all geographies Europe is probably the one that's most dependent on energy and we need to drive the energy transformation more quickly than other geographies.

00:31:45: That are less dependant on external supply of energy.

00:31:49: And when I say energy always mean at end will be generating electrical energy from whatever original source.

00:31:59: Thank you Peter.

00:32:02: Let's change to the human side, The workforce.

00:32:07: as AI becomes embedded in energy systems and industrial operations As you described it.

00:32:14: what skills will the automation engineers?

00:32:17: And energy professionals need so that they Can use it and which skills which they don't have today typically?

00:32:27: So let me first answer this question on a bit of higher level before I go into the specific requirements for energy and automation engineers.

00:32:40: What we are seeing happening is the most rapid evolution of technology that I've seen in my career.

00:32:48: And this is something that excites me, and it's a very interesting phase for professional careers so far because technologies are moving fast.

00:33:00: sometimes i also get worried.

00:33:02: if you only look at last six months You can see an evolution in AI technology which becomes extremely transformative And probably the most visible area is software engineering today.

00:33:19: While software engineer was a skill that kind of challenging to find best people in market, this has totally changed now.

00:33:35: business processes will change dramatically.

00:33:38: So most of the coding, most of generation test cases and testing will be done by AI.

00:33:44: Humans have more role to oversight in creating the right architecture.

00:33:50: Now this changes an ecosystem of software engineering pretty significantly.

00:33:57: I came up with a conclusion that that this is not only happening in software engineering, but it's happening for all of knowledge work.

00:34:10: And knowledge work is in finance and sales and marketing and HR.

00:34:14: so I expect a transformation which is bigger than the internet transformation thirty years ago... ...and its moving at faster pace Now.

00:34:25: what does that mean to our workforce?

00:34:27: What do you think about skills?

00:34:30: Certainly It's fair to say that not total jobs, but tasks of jobs will be replaced by AI.

00:34:39: And companies probably better start getting their arms around there to understand which are my jobs?

00:34:45: Which functions will be impacted by AI and how much?

00:34:51: What does this mean for our

00:34:53: workforce?".

00:34:54: There might different conclusions... ...but I'm pretty sure everybody's role will change.

00:35:02: There will be certain tasks in my role that we'll done by AI tomorrow, there are other task only emerge through AI right?

00:35:11: So I'm dealing with things that AI can't do yet or AI create.

00:35:16: some other challenges need to be dealt but every company has the responsibility looking into this and understand what is impact of AI on on the workforce.

00:35:29: We are doing this very actively at Schneider and going now a bit more specific into your questions, while I believe that any knowledge worker will change through the impact of AI over time in some areas pretty quickly.

00:35:47: Also i'm deeply convinced for our automation and energy engineers their role.

00:35:56: By the way, not everybody has to become a developer of AI products and algorithms or LLMs.

00:36:02: But everybody will be using these right?

00:36:04: So so everybody needs to think about companies need.

00:36:09: How do we have to upskill our workforce that they can use all these AI capabilities in the best possible way?

00:36:18: And I don't think there is any way around upskilling and continuous learning on this.

00:36:24: We need to embrace this transformation very quickly because it's moving at rapid speed.

00:36:34: We need to go into a more strategic workforce planning, we need to get into clear upskilling learning and training capability.

00:36:42: And everybody –and I really mean EVERYBODY– will need to learn how the AI capabilities are in as best possible way.

00:36:53: So Peter if you take a closer look at our calendars schedule follow-up in twenty thirty five and polish out our crystal ball.

00:37:02: What development at the intersection of AI and energy would probably surprise people the

00:37:07: most?

00:37:09: Look, I think in thirty-five which is roughly ten years from now.

00:37:16: I think what we have been speaking about how do you make the grid more intelligent... ...how to provide renewable energy with more sustainability and management?

00:37:31: I would hope that this all happens in the background, right?

00:37:37: So with it we have built out our fleet of renewable energy generation.

00:37:43: That we've build-out our battery capacities and then... ...that we have filled up our grid and made our grid fully intelligent and autonomous.

00:37:53: so.. ..this is happening without much human intervention.

00:38:00: And let's get to a bit more personal question What motivates you, Peter?

00:38:05: Getting up every morning driving change at the intersection of the old technology energy transmission and distribution on one hand side.

00:38:12: And then your technology AI.

00:38:14: what is it?

00:38:16: look I find that I said a few minutes before The most exciting phase off my personal and professional career.

00:38:26: I had the opportunity to live through the internet transformation, i still remember my first email or website.

00:38:35: that went too right?

00:38:36: So in this group that grew up before the Internet if we go thirty years back and compare these two today.

00:38:51: We couldn't do anything without the internet anymore, right?

00:38:54: we couldn't record this podcast.

00:38:57: We could not watch a movie from last night or shop what we were shopping over that weekend and find our way to next place where you want to go.

00:39:09: so the Internet has influenced all of your personal lives as well.

00:39:18: I think AI will be exactly the same But probably moving at an even faster speed.

00:39:27: And I let you judge if this is good or a bad thing, right?

00:39:29: I find it definitely exciting that technology going to transform our private and also professional lives.

00:39:42: Now why do i think particularly fascinating in the role that im in?

00:39:47: Of course, AI is the foundational technology that's driving that change.

00:39:52: But I hope it became clear from our discussion that AI and energy are really connected by the hip.

00:39:59: There is no AI without reliable and sufficient and sustainable energy supply And on the other side there will not be an energy significant support of AI, of controlling the energy flows.

00:40:16: So two technologies connected by the hip and I find it totally exciting to be part of that ecosystem where energy companies in AI companies work hand-in-hand to deliver their next foundational technologies.

00:40:33: Okay i feel your enthusiasm which is a good feeling Peter.

00:40:40: are there any topics we have not discussed yet?

00:40:43: which matter for you and might matter to the audience?

00:40:47: Look, I think we covered a lot of ground over the last forty minutes.

00:40:52: And if i have only two messages or nothing new that we haven't covered then it's probably two main messages that want to underline.

00:41:04: message number one is AI is a transformational technology.

00:41:08: they will impact all us every Every industry, every company.

00:41:13: Every person will impact our private lives and it's very much coupled with the energy transformation right?

00:41:21: So these two things go hand in hand.

00:41:23: that probably message number one.

00:41:26: Message Number Two is this a real transformation?

00:41:32: um This Is A Transformation That Will Have An Impact On Our Jobs on Our Tasks.

00:41:39: It will have an impact on our private lives, right?

00:41:42: And there are two options.

00:41:43: You can fight that transformation and most likely we'll not win or you better embrace it and adopt apply and transform wherever makes business sense.

00:42:03: Peter with this we've reached the end of today's episode.

00:42:06: typically guests for three key messages.

00:42:11: You addressed it all in a very fine way, if there's nothing more to add I would like thank you so much!

00:42:17: It was great pleasure having as an international guest at our podcast.

00:42:22: Thank-you so much and what is the pleasure of me?

00:42:24: Thanks for being with us today.