Sophia Mendelsohn: [00:00:00] AI's capabilities are entirely inseparable from data centres. And data centres are inseparable from land, water, power, waste, heat at scale. Most credible projections assume demand will keep rising. Demand for electricity, demand for those precious natural resources will keep rising. So framing AI as sort of just software allows leaders to defer the hard infrastructure question a little bit. And that deferral is a governance failure.
Tom Raftery: That's the crux of this conversation. AI's promise is real, but the choices around it are already becoming business, climate, and governance decisions. Good morning, good afternoon, or good evening, [00:01:00] where everywhere in the world, this is Climate Confident Stories and Strategies that cut Emissions Episode 284, and I'm your host, Tom Raftery.
My guest today is Sophia Mendelsohn, who leads SAP's Global Sustainability Platform. We talk about how sustainability teams can move beyond PDFs and carbon accounting and into procurement, supplier data, financial planning, and the decisions that shape AI before those systems lock in.
So I began by asking Sophia how sustainability and commercial growth come together. And very quickly the conversation moved to the infrastructure reality behind AI.
Sophia, welcome to the podcast.
Would you like to introduce yourself?
Sophia Mendelsohn: Yeah. Thank you Tom. Thank you to everyone giving us their time to listen. I am Sophia. I run SAP's Global Sustainability [00:02:00] Platform, including corporate as well as the products and solutions that help others also execute their sustainability.
Tom Raftery: Interesting. And so you sit across both sustainability and commercial growth. How do those two worlds come together for you?
Sophia Mendelsohn: Well, they come together through ensuring that we know exactly what needs to be done in the world of sustainability. In technology, there is a philosophy that the closer you are to the problem, the closer you are to the people and teams executing on that problem, the better off the solution.
Tom Raftery: And was there a moment for you when these two worlds stopped feeling separate?
Sophia Mendelsohn: Oh, these worlds were never separate to me. I've been a Chief Sustainability Officer three times. I've been in this space over 20 years. [00:03:00] I know that the major challenge for sustainability teams are putting those precious time and resources from the wrong space to the right space. Sustainability teams don't get outta bed in the the morning gung ho to create PDFs.
They're not excited, right? About counting things that they don't see the world valuing. No. We want to use technology and automation to build up the governance, the infrastructure, the accounting, the carbon accounting elements in particular as quickly as possible, and free the sustainability teams to get to the change management work.
Tom Raftery: And let's get into the tension because this is where things get interesting. AI, AI is not a weightless. So what are most leaders still getting wrong about AI and sustainability?
Sophia Mendelsohn: Yeah. And you're [00:04:00] right, Tom. Tension is where things get interesting, right? And sustainability professionals, and those in the world of climate solutions sit at attention, have volunteered, have put their hand up to be part of that tension. And the same goes for AI, right?
So first and foremost, let's establish two basic facts. One, AI's capabilities are entirely inseparable from data centres. And data centres are inseparable from land, water, power, waste, heat at scale. So most credible projections assume demand will keep rising. Demand for electricity, demand for those precious natural resources will keep rising. These are engineering constraints within real geographic regulatory and community dimensions, In SAP's own integrated report in 2025, we acknowledge that the growing demand [00:05:00] for energy due due to increasing use of AI could have a negative effect on GHG emissions. We see the same from the World Economic Forum, but the upside is also real.
So is the downside. So framing AI as sort of just software allows leaders to defer the hard infrastructure question a little bit. And that deferral is a governance failure, right? So if we're going to talk about AI, we're also gonna have to talk about sustainability.
Tom Raftery: Okay. And obviously as you said, people think that AI is just software, but what changes when we treat it as physical infrastructure?
Sophia Mendelsohn: Mm.
Tom when we see AI as physical infrastructure, right? The data centre comes in to focus, and this is structurally identical to [00:06:00] what mining and energy companies have navigated for centuries. The discipline built to manage all that contradiction, all that tension that we talked about earlier is called sustainability.
The difference is that the technology centre is arriving at this realisation now, right? As this technology builds and grows and brings benefits to us, whereas mill and mining and other traditional energy industries had a hundred, a hundred and fifty plus years to deal with it. This is like watching history in fast forward.
Tom Raftery: Right. Well then where do you see the real business stakes if companies get this wrong?
Sophia Mendelsohn: There are real business advantages at stake AI can bring, and it has already brung, immense benefit and changes to companies. There is no company that's gonna leave [00:07:00] that advantage on the table. So we are all opting in to proactive, conscious, additional resource consumption. And if we seek only the benefits without counterbalancing the reality of investment for those benefits, what supports the business case will eventually falter.
We will eventually come full circle to the need to say, we want the benefits of AI. That means investing in understanding the unwanted externalities as well as the resource consumption going into it, and investing in building up the accounting and governance infrastructure to support that.
Tom Raftery: And then when we think about reporting and decision control, when does sustainability stop being a reporting exercise? Those PDFs you referred to and become a [00:08:00] decision rule inside the business?
Sophia Mendelsohn: Mm, today, please, Tom, let it be today. There's a historical perception of sustainability as a reporting function. And that is a liability at this moment. Sustainability, friends do not approach AI the way we approached ESG reporting, We do not want to be bean counters of carbon and water associated with AI while missing the largest technical opportunity to advance a balanced economy that we've seen in our lifetime.
And I'll, I'll make a pitch directly to my fellow sustainability leaders here, right? You bring three capabilities that are necessary for your organisation to scale and grow and benefit from AI. So, you know, one, how to manage resource constraint management. [00:09:00] Two, you know how to do multi-stakeholder engagement.
And three, you understand how to administer the social licence to operate.
Tom Raftery: To what end?
Sophia Mendelsohn: To the end of being able to use AI within planetary boundaries, within social acceptance, and within the benefit of your own company's understanding of the ROI of the technology. Right? So I think, Tom, we're already seeing this a little bit, and you know, podcasts live on far past the moment they're recorded.
But if we take ourselves into this moment for a second, with every technology, with every trend, there's a hype cycle, right? We saw it with sustainability as well. And within the upward initial curve of that hype cycle, there's a temptation to disregard any [00:10:00] consequence of the new topic, to not consider the pros and cons, the different stakeholder groups.
And without that consideration, you get what Gartner, which is a major analyst in the technology industry, refers to as a hype curve or a hype cycle that then comes crashing down much faster. And organisations, companies, they do best with sort of slow, stable improvements that then continue in an upward trajectory.
And I would argue as a corporate professional, any sustainability executive or team member who's opted in to a company structure, we have a fiduciary responsibility to help our organisations not just get the fastest return on AI as a technology, but the most consistent and sustained. And that's what we're seeing happening right now.
When you're deep in it like I am, you see [00:11:00] organisations realise like, Hey man, tokens actually cost money and maybe we should hold the outcomes of AI to any other investment we would make. Show me the benefit to my people. Show me the benefit to my business.
How do I create a quantifiable outcome for this? That's the side of the equation sustainability wants to be on. What are the benefits to materials R&D? What are the benefits to new technology that do help us live in planetary boundaries? You do not wanna be on the side of the equation of tallying up the bill and then going, oh gee whiz, we don't know how to solve it, so use less of it.
Tom Raftery: And obviously then we gotta get into the issue, the thorny issue of data. Because if AI is making recommendations from [00:12:00] fragmented sustainability data, what goes wrong? Are we likely to be making bad decisions just more efficiently?
Sophia Mendelsohn: Well, as has been said many times before, AI is a reflection of ourselves. It's a reflection of what we've done. It's a reflection of how we train it. And what's clear in the business world is that we have an inherent bias towards the short term, very understandably. So if we train AI to do business case analysis, for example, as seeing sustainability as inherently, moral or inherently for a greater good outside of our business structure, and AI is inherently within the business structure because it creates efficiency. You'll get the wrong answer.
And it's worth, Tom, if I may, before you ask me another question, just pause here for one second, folks, because the sustainability [00:13:00] elements we're talking about here are very short term. Community backlash against data centres holding up hundreds of billions of dollars of investment that is real today. Concern about energy and water use in local communities and social licence to buy in onto data centres real today. And likewise, if you think you can create a beneficial business case or a superior analysis for an organisation without considering the changes in natural resources, without considering supply chain, you'd be grossly mistaken.
It would be like trying to buy olive oil without seeing what olive oil and olive farmers are saying in Greece. Or it would be like looking at cacao without noticing the trends of cacao prices and droughts in Africa.
Tom Raftery: Mm-hmm. As a guy based in the south of Spain, [00:14:00] in the middle of olive country, I do have to point out that Spain is the world's largest exporter of olive oil, but the Italians get all the credit.
Sophia Mendelsohn: That's okay. And so, so Tom, if I may ask you a question, right? You sit at a country already feeling the effects of climate change, I think with, a proud history, of things like olive oil. And is there a global, or social, I should say, is there a community-wide dialogue about the effects of olive oil in Spain?
Tom Raftery: I mean, there is, but only because it's massively consumed here, as well as produced here. So if you remember, two years ago, two summers ago, so summer 24, the year before that, again, 23, we had a significant drought here. So the olive oil harvest dropped by about 40%. And so the price two years ago, [00:15:00] so six months after, the price in summer 24 of olive oil went from about, five euros a litre to between 10 and 12 euros, a litre for extra virgin olive oil. So in that scenario, yes, it is very, very close to people's, attention span, but it's gone back down now again. It's, it's at about five again per litre. Now it's quite variable, and so we'll have to see what it's like next year after this summer because, and, and this summer, even after, because with the El Nino effect coming in, there's a large possibility that we'll have a, a drought again next year, next summer.
So the year after the, the harvest may be impacted again. We're quite lucky this year because we had a very wet winter. And so, the olive harvest should be good this year.
Sophia Mendelsohn: And so that's, that's exactly the question. I, is that a sustainable way for an organisation to go on? As sustainability professionals, when we see that our [00:16:00] communities or companies are exposed to price fluctuation at this rate, is our job to say, this is how much emissions came out of an AI data centre, or is our job to say, this is how much emissions came out of a data centre.
Now let's make dang sure, we reduce past that through additional benefits. I don't mean just direct carbon capture offsets, which are an important part of the equation, but I mean in the ability for sustainability professionals to use AI to talk about baseline and shifting baseline, right. We are also in a moment in time as this podcast is happening right now where gas prices all over the world are going up and down in geopolitics, and I'm constantly asked, oh, is this it?
Is this where sustainability is coming back? Is this where we see the return in the pendulum swing? And I'm like, heck no. These are short term things. Olive oil prices went [00:17:00] back down, gas prices will go back down. It is about pointing out the sustained disadvantage, the sustained volatility, the sustainability professional, using AI to understand and quantify that risk for their organisation so that we can use AI to create a living intelligent institutional memory for the decision makers in our companies that goes beyond what olive oil or gas costs today.
Tom Raftery: Sure. Sure, sure. And in terms of the, social licence that you talk about, is the constraint more likely to come from communities, or regulators, or customers, or employees, or someone else entirely, or all of the above, or it depends?
Sophia Mendelsohn: So let's start with what social [00:18:00] licence to operate even is. Social licence to operate is sort of the informal permissions granted by communities, customers, and workforces for a company to operate. It's different than legal compliance. So right now for AI, we see this popping into three places. One, the communities hosting the infrastructure.
Two, the enterprise customers whose own stakeholders scrutinise their technology choices. And three employees who must trust the systems they're being asked to use daily.
Tom Raftery: And obviously we're seeing a lot of companies racing into AI, but governance, as you referred to earlier is lagging. So why is accountability so hard to assign?
Sophia Mendelsohn: Why is accountability so hard to assign? It's so true.
Accountability for ethical AI and AI and [00:19:00] sustainability challenges our conventional organisational structures. It challenges the typical boxes we like to put people and ideas in, So AI typically sits in technology and now we need it to sit alongside technology, people, and environment. There's no ability to say we want to work on re reducing emissions, but we're not gonna reflect on perception of data centres.
Tom Raftery: And what's the cost of waiting until after deployment to govern AI properly?
Sophia Mendelsohn: The cost of waiting for sustainability to engage in AI will be the missed opportunity, This is our window of opportunity to say how you design a [00:20:00] chip for water consumption matters. This is our opportunity to say, no, I do care which data centre my resources for my AI use cases and my token consumption are being pulled from.
And I would like to know that that data centre is covered by renewable energy, either in the local market or through some type of market based measure. This is when whole new worlds of procurement are taking shape. And I would implore my sustainability coworkers do not use this moment debating or reporting out on how much AI emissions increase your scope 3. Yeah do that. To be clear, Tom, for the comments, do that, please do that because we are the accountants and record holders. Do [00:21:00] that, but also go to procurement. Go to your global risk function and say, we are now signing new contracts for the first time. We have new companies new types of companies asking us to engage with them for the first time.
Now is the time to bring sustainability to the negotiating table and say, if you want us to go with your data centre, if you want us to go with your forward deployed engineering teams, talk to us about how you are gonna control for water, land, and energy consumption. So I know my brand is not exposed. I know my Scope 3 isn't exposed, and I know you have a way to control costs in the long term.
Today is the period to ask questions because the system will solidify itself. It will calcify into something that scales and the [00:22:00] messages that the market gives now to those organising and decision makers are what will be entrenched in the larger governance system for the years to come.
Tom Raftery: So what you're saying there is that sustainability teams are not best placed to solve this alone, they need a broader coalition.
Sophia Mendelsohn: Oh, good Lord. I've never seen anyone solve anything alone. And when it comes to AI, while sort of the act of coding or vibe coding can be very individual, the act of applying the outcomes of that coding to society is perhaps more social and communal than anything we've seen in recent times.
Tom Raftery: And where have you changed your mind on responsible AI or sustainability over the last year?
Sophia Mendelsohn: Oh, Tom, it's a hell of [00:23:00] a question. where have I changed my mind on sustainability and AI over the past year? I have gone from thinking that we could look at energy consumption and direct carbon air capture and market-based measures alone to really seeing AI as a gift to sustainability teams because we now also have a tool to calculate externalities. We now also have a tool to create the assumptions and data sets that we need. And let's just take Scope 3 for an instance. Okay. I'm assuming everyone listening to something like this knows about Scope 3.
Scope 3 is based on the belief that if you ask, you will get information back. That is not correct, right? We know what it looks like inside a supply chain. You could ask and not get an answer. By the way, [00:24:00] shout out to the sustainability community on LinkedIn for their hilarious gifs, and JPEGs on Scope 3 emissions and asking for suppliers for information and not getting responses back.
But with AI, there's the ability to generate that information for yourself and say, dear supplier, is this right or wrong? And you have the opportunity to prove it to me otherwise, but I'm gonna go forward and say, this is your carbon footprint, this is your product carbon footprint, and I'm expecting you to minimise that.
I would also say, let's talk about sustainability, AI and investors. Okay, so slightly different headspace, gear change for our conversation, but an important one. I spend time with investor relations. I spend time with investors. And what AI is making clear in that space is that investors have [00:25:00] the opportunity to model scenario plans and different potential outcomes as much, if not more, than the company and the portfolio themselves. So this is a huge cue for sustainability professionals, Do not remain the asker for information. Transition to becoming the calculator of the information with AI in your own hands, and then pressure testing if that works with your stakeholders.
But establishing for your own organisation, a net new amount of data, a net new baseline. And Tom, that baseline is so important because baselines within human psychology and human nature are shown to shift about every two years. So I'm [00:26:00] sitting outside right now. It's hotter than it historically has been, but I'm not really processing that because it's been hotter than historical averages in the summer where I am for the past two years. My baseline has already shifted. Actually even now, socially acceptable olive oil prices are probably higher than they would be without climate change. But the Spanish baseline for what is acceptable to pay for olive oil has already shifted. Right?
And then we look for all these other ways to compensate for that, ways to cure the symptoms rather than the root cause. And if sustainability professionals are able to use AI to scenario plan what their company's supply chain looks like in the foreground, or go back to the background and say, Hey, that supplier that went out of business, or that spike we saw in automotive parts that was related to historical flooding in Thailand X number of years ago.
And by the way, those [00:27:00] prices came back down, but they never came back down to the same baseline before that climate crisis moment. Right now we're having a conversation.
Tom Raftery: Sure. Sure, sure. And so then what does it mean to build sustainability into AI decisions as opposed to bolting it on afterwards?
Sophia Mendelsohn: To build sustainability into AI decisions rather than bolt on after, means to consider the entire equation of AI and sustainability as a return on investment. We are investing a certain amount of human capital, land use, water use, and energy use into a net new technology. What is the return on investment for your organisation, for the community you're from and for the long term objective you might be trying to achieve? In this case, I would say a low carbon circular economy that can still grow and thrive and support more and more people.
And [00:28:00] right now, dear sustainability professionals in your organisation, that calculation is being taken at the board level with the CEO level. And the return, or the outcomes are being listed. We get through this workflow faster. We eliminate this cost. We get this type of information that helps us grow in new markets.
We increase margin because of AI,
Tom Raftery: Hmm.
Sophia Mendelsohn: And our job is to make sure that sustainability outcomes are added to those returns, right? So the investment is being made already. The dollars are being spent. Now, our job is not to count the impact of the dollars spent alone. It is also to make sure that outcomes include the objectives of the [00:29:00] sustainability team within that organisation.
And then once you're in that return on investment conversation, and you're looking not just at, for example, margin, decrease through more efficient processes in the factories, but also the ability to use AI to rethink R&D for packaging, and reduction in packaging fines, fees, and just sheer amount of plastic used. Now, you're part of the benefits conversation for the business.
And you're earning your seat at the table by articulating an outcome that your organisation should receive for investing in AI and changing the balance of equation. This much money spent, this many more benefits, and that is a much stronger place to then go ahead and continue to address land use, emissions, and water from, rather than being the bean [00:30:00] counter or the one trying to stop or pull back on a technology per se.
Tom Raftery: And can you give me, one practical decision that AI could improve if sustainability data were properly included?
Sophia Mendelsohn: AI plus sustainability can bring real benefits immediately. Let's imagine you are a beauty company, right? And you're selling perfume. Perfumes are actually made of natural ingredients, very beautiful natural elements like vanilla and rose petals, These are precious inputs to your commercial product.
In the past, a corporate sustainability professional would've had to pay sort of X dollars to a consultant to go through a complicated scenario analysis per raw ingredient, and then get it back as a PowerPoint [00:31:00] to try and pitch to the product team about why they should care about how hot it is where you grow rose petals.
Now for every input into that perfume, you can run a scenario of how endangered that ingredient is, what could be done with local communities to protect that commercial input and cost to your commercial product, how you might need to pay more for shade grown ingredients rather than regular, and how you might need to compensate for that cost by changing the packaging and turning around and going to the packaging manager and saying, let's rethink our R&D on this.
Let's see if we can get plastic out or switch from glass to plastic in a different type of example to bring cost down and then go to the marketing department and say, dear marketing, [00:32:00] you've always assumed that the driving force of this perfume was how it looked on the shelf. Now let's test that, right?
We can use AI to do many, much more virtual outreach and prove that actually maybe there's more customer engagement in changing the packaging, for example, to recycled packaging or a material made of plastic harvested and saved and pulled out of the oceans. And there's an engagement story there as beauty companies I've spoken to have seen as sneaker companies I've spoken to have seen.
And it allows the sustainability professional to go department by department in the value chain of whatever their organisation sells and say, let me help you rethink our assumptions, because now we have the budget and bandwidth to do it.
Tom Raftery: Nice. And what changes then when carbon [00:33:00] moves from sustainability report into financial planning?
Sophia Mendelsohn: When carbon moves from your PDF report to your true financial plan, the first thing you do is make it approachable for your colleagues, right?
They're not thinking in scopes, they're thinking at the group level, they're thinking at the enterprise level, the country level, the P&L level, the portfolio level, and ultimately the SKU or product level. So if you are going to approach someone with an ask, a desire for investment and change, we gotta be speaking the same language as them.
And because it's you with the ask, we're gonna speak their language? After it becomes approachable, you can make it relevant. What does this have to do with me? What does this have to do with my iPhone instead of with Apple and the memory of Steve Jobs? And then you can begin to scenario plan for it, right? More and more of the world has a [00:34:00] price on carbon.
We see the cost and polluter plays principles being distributed to companies as voters and their elected officials are reluctant to put on general costs in a high inflationary environment. And companies need to be able to account for what is in their product and negotiate with the regulators or the equivalent of the tax man and say, this is how much we have in there and this is how much we don't.
And you need that topic to have been approachable and relevant to your colleagues first, if you're then gonna do that.
Tom Raftery: And can you turn this whole argument into one sentence a CEO would remember?
Sophia Mendelsohn: Well, Tom, if I had to give a CEO one sentence to remember, I would start it with a compliment.
I would say [00:35:00] as the most intelligent and experienced person in the room. You have been through waves of excitement about change in technology before, and you know that the organisations that get the most sustained benefit from that change are the ones that bring multiple stakeholders in from across the company to determine the best outcomes for themselves, and then blend and prioritise them for investment.
Having sustainability in that blend of outcomes not only gets you an additional outcome, it protects all the others.
Tom Raftery: Lovely.
So can you take this, Sophia, from where we are at kind of ground level now to going up 3000, 5,000 a hundred thousand feet. [00:36:00] Give us your overview of it.
Sophia Mendelsohn: Sure. So in the immediate right at the 30,000 foot level, this is headed towards the sustainability team, making sure they are canvassing their organisation to understand who is spending what on AI and determining how sustainability could be an additional advantage of that. Let me give you one really clear example on that, Tom, for the short term, right?
We have one company in the technology space, the consumer goods technology space, using AI to better answer questions they receive from their clients as a vendor, and the vendor questions they have. At first, they were just answering questions about price. When we engaged the sustainability team with them, they were able to quickly and automatically calculate thousands of product carbon footprints across multiple [00:37:00] locations.
Each location of course, having its own special different process and say, this is the price and these are the additional benefits and attributes we give you at the product level of what you're asking for. So the ability to generate an additional return on investment, an additional listed outcome in the short term. And then in the medium term, there's the protection of the resources as we've discussed many times. There's also the protection of the company, Hey company, you're signing up for a lot of money. That's a lot of spend. Company, we should think about what we could ask for while engaged in that spend, or what additional costs to that spend might be.
And then at the highest level, that a hundred thousand foot level, the sustainability professionals should get excited how much data [00:38:00] and points of view this will be able to calculate for them on their existing budget and what type of tools that can literally put in their hands to go back to their stakeholders and say, your baseline is shifting and you don't even know it, but I got the data to help you track that now.
And I have views on where it's going.
Tom Raftery: Okay, great. For people who are listening to this, Sophia, who want to act on this, where should they start?
Sophia Mendelsohn: Well, Tom, I'll, I'll answer that question as a leader, What do I hope my team members come to me with from AI on where to get started, my customers come to me on where to get started, and our regulators, or elected officials and other stakeholders and influencers like the media. So for the team member, we want the team member to get started with AI by canvassing their organisation, understanding where it's in [00:39:00] use for what purpose, and determining if there can be an additional outcome related to sustainability or additional protection added to that. And using a balanced equation, using AI tools to make a data-driven argument that used to cost hundreds of thousands of dollars of consultants and now doesn't. And that's something my own team's done and that I've seen for myself.
Tom Raftery: Okay.
Sophia Mendelsohn: For my customers, we're seeing them say, Hey, I didn't have this information from my supply chain before, and even when I did get it, I didn't believe it.
But now with your tools, I'm able to calculate it for myself and work from that place rather than waiting for the world to come up.
And for stakeholders, be them media or influence or regulators to understand that AI is an infrastructure question and a social buy-in question as much as it is a technology [00:40:00] question.
Tom Raftery: A left field question for you, Sophia. If you could have any person or character, alive or dead, real or fictional as a champion for sustainable AI at scale, who it be and why?
Sophia Mendelsohn: Well, I am sitting in a field, and if I look to my left and I had to imagine a historical figure coming or a fictional figure, I would choose the Greek goddess Athena, because Athena knew that you had to get in the mix if you wanted to apply your wisdom. In every story, both Athena and the sustainability professional refuse to be passive accountants of the change, but determine how it should be designed [00:41:00] for what outcome, at what cost.
Tom Raftery: Very good. Lovely. I like it.
So, Sophia, if people would like to know more about yourself or any of the things we discussed on the podcast today, where would you have me direct them?
Sophia Mendelsohn: Oh SAP. Sustainability. I used to give a website address, but you know, now you can just put it into your agent. Tom, for anyone who wants to continue the dialogue and go deeper, I am engaging on LinkedIn on exactly this topic.
Tom Raftery: Great. I'll put your LinkedIn details in the show notes and everyone will have
access to them.
Sophia Mendelsohn: Thanks.
Tom Raftery: Sophia, that's been fascinating. Thanks somebody for coming on the podcast
All market data is provided by Barchart Solutions. Futures: at least 10 minute delayed. Information is provided "as is" and solely for informational purposes, not for trading purposes or advice.
To see all exchange delays and terms of use, please see disclaimer.
