Latest / Real Estate Meets AI / A Conversation with Alex Wolkomir from McKinsey & Company
Transcript
- Sandeep Ahuja: Here's what fascinates me about consultants who advise on AI and real estate. They see everything, every company's playbook, every failed pilot, every breakthrough nobody's talking about yet. Today, I'm talking with Alex Volcomere, a partner at McKinsey who co-authored the report that put a number on what Gen. AI could unlock for our industries. We're talking close to $180 billion. Here's also the guy who recently told a room full of real estate executives that use cases are actually the wrong way to do AI. So let's dive right into it. Alex, welcome to Real Estate Meets AI. A quick intro for folks who don't know you. You're a partner at McKinsey and Company based in New York, leading the firm's work at the intersection of AI, digital strategy, and real estate globally. Before McKinsey, you were at JPMorgan Oliver Wyman. You did masters from Stanford in design and engineering. Oh, and somewhere. In there, you went to to work on innovation for refugees with the International Rescue Committee during the Syrian crisis. That is ⁓ a typical consulting resume. But here's what I actually want to know. You describe yourself as a food junkie. So if you had to cook one dish to win over a skeptical real estate CEO who doesn't believe in AI, what are you making? Alex Wolkomir: an interesting I feel like it has to be hearty simple. Probably like a braised short rib with mashed potatoes. something feel like I'd made a bunch of times, of have that pattern down. I think there's something yeah, just the ⁓ simplicity, the heartiness, and it kind of gets to the core real good So, yeah. ⁓ Sandeep Ahuja: I love that. Real good raw flavors. ⁓ I can almost feel it. Alex Wolkomir: Yeah, it's back down to the basics. Yeah. guess in some ways similar to I think where we're kind of at with AI, which is like getting to the basics, getting to the workflows and to what matters versus all the hype, just really leaning into the raw simplicity of things. Sandeep Ahuja: I love that you say that. So obviously you've been advising the biggest real estate firms in the world on this stuff. You've seen the hypes, you've seen the pilots, you've seen the wins. So we'll kick us off with what we call the AI meter on a scale of one to five, where one is AI-averse and five is AI-forward. Where do you think the industry is at right now when it comes to AI? And tell me why you picked that number. Alex Wolkomir: I would probably say... On the whole, the industry is probably at around a two and a half to a three of five. And I would say there are some players folks in the industry that ⁓ are closer to a four in like ⁓ pockets. I there are a lot of teams out there that are experimenting, trying different pilots, doing different things. know, they're summarizing leases, drafting marketing comms, they're thinking about reporting and how make sure all the data is really clean. some folks that are really pushing it that step further of having AI actually run in the workflows, taking action. And we're kind of moving out of this world of I'm going to give it a prompt and ask AI something to actual action and doing with the right guardrails, with the right governance. So I'd say broadly, a lot of folks experimenting, a lot of piloting, a lot of really exciting companies in the world pushing the boundaries. And I'd say a few folks that are it right and reaching for that four in my mind. Sandeep Ahuja: That's really exciting. So for everyone that's listening in, at least this is what I hear. You're not behind. Everyone's somewhere in the middle there. Most folks closer to that two and a half, three, three-ish mark. So if you feel like you're behind, you're probably not. You might be right there in the middle somewhere. So long as you're doing something with it, is at least what I'm hearing. Alex Wolkomir: That's absolutely right. mean, big part of it is just the experimenting, the learning, the exposure, if your point's into Sandeep Ahuja: Love that. So let's get into it. Obviously, you co-authored McKinsey's landmark report on gen AI in real estate. one that said that technology could unlock 110 to 180 billion dollars of value the industry. Now that was late 2023. We're now in 2026. Where we on that journey? Has the reality matched what we projected? Are we over it? Are we under it? Like, where are we? Alex Wolkomir: coincidentally, we actually have our next version of that report coming out early next month. So a good lead in that we're actually on the precipice of some new numbers. when we read it, a lot of that analysis we looked across, I think it was 48 countries with like the McKinsey Global Institute, which is McKinsey's kind of sister or think tank. looking at labor productivity, looking at, as we think about of the new things AI is around agentic systems that may have just been ideas back in 2023. When we look across real estate and construction, we think globally the opportunity is actually even bigger now. we're more in the realm of like 400 to 550 billion, if you think globally. So I actually think are we is that it was true and in many ways we were conservative. Cause I was kind of thinking about generative AI, the use cases of the moment in 2023. Sandeep Ahuja: ⁓ my. Alex Wolkomir: we're very much of that mindset of help understand, ask a kind of go in that mindset. And I think in a very different world now sitting here in 2026 where it's really around help me get it and let AI ⁓ act. So I think anything, ⁓ we're in an exciting new where there's a lot more possibility. And I think we're starting to learn where there are pockets where AI can really be transformative in terms how real estate is doing business. as usual. Sandeep Ahuja: I love that and it's so exciting to hear that it's so much greater than even you and the team that put that report together anticipated when you did that work. Do you think we're gonna be here again in the next three years or do you think we'll be there faster? we were like, ⁓ my gosh, it's actually, you maybe 1.5 trillion. We significantly underestimated it yet again. What do you think? Alex Wolkomir: the opportunity may get bigger. And there's also just natural economic growth. I think the real question is that's the number of value is available. And it's really going to be how do folks actually operationalize move to capture it. Because that number, you know. in some ways is the target. If everybody were to go out and adopt it in the ideal way, the value. But I think to your prior question with the AI scale, folks are very different places. ⁓ So there's an acceleration pace. There's of the natural forces of the industry that will determine how quickly we get But I would say I am personally optimistic that I feel like moving in right direction. And there's a buzz and an energy in the industry now around AI that Sandeep Ahuja: Mm. Yeah. Alex Wolkomir: even two years ago, I think it's been step change in terms of what I have been feeling and hearing. I mean, I was at the International Builders Conference down in Orlando last week and AI was everywhere. It was all anybody was talking about amongst, there were 70, 80,000 folks down there. But it's palpable in all corners of the industry right Sandeep Ahuja: Yeah, yeah, that's definitely what we're seeing too. But it's all about not just talking about it. What you said earlier on resonated so closely with me. It's not about the hype. It's not about buzzwords. It's not about can you use chat and GPT or Claude or something, you know, like it's like, well, what is it meaningfully doing for your business? And that's, think the thing that most real estate companies and developers that we talk to, which we talk to hundreds every, every week. That's what they're focused on. They're like, good, you're AI powered architecture firm. What does that even mean for me? I don't want to do AI just for the heck of it. Alex Wolkomir: Yes. an outcome is not ⁓ are using an AI tool. or minutes logged in or folks logging in daily, what are the actual business outcomes? Like the lease execution happening faster? Are we our ⁓ maintenance and orders more efficiently, adhering to SLAs in a construction Are we reducing the number of days to delivery and not having mismatched vendor scheduling in of when plumber shows up versus the carpenter for things that have to be there's a little bit at times I an organizations of they get fixated on means people use AI versus real KPIs are what's actually value for the company you think about those end state metrics that matter. Sandeep Ahuja: Mm-hmm. Most definitely, I couldn't agree more. Now, obviously in this report, which it sounds like is soon to be an outdated version of the report because there's a new one coming out in April. for April, so for all that are listening in, definitely March. So everyone that's listening in, go check out that report. But in the original one, you laid out the two by two framework, right? Like the two quick wins plus two aspirational use cases. ⁓ Alex Wolkomir: Hahaha March, very soon. Sandeep Ahuja: Give me a real example. What's a quick win that's actually paying off for firms that you're advising? You kind of touched on it with real metrics, but maybe a little bit more. Alex Wolkomir: Yeah. I'd say two really wins that we're seeing. I'll give you two examples. One think, How do you have, whether it's the sales office or the leasing office, on 24-7 and just ⁓ kind be like voicemail that responds, but actually do things in the systems that matter. Book tours match folks to properties that may fit them in life stage from a segmentation perspective. Qualify leads and help figure out where they should go, but have that actual ⁓ interaction. But I think is a ton hitting that top of line in terms of do you kind of meet people where they are. Because a lot of times I'm searching for home for my apartment that might be happening, you ⁓ 10pm night browsing around on my phone, having these and the chances are like someone's not there at 10pm to answer my question about, you know, financial or answer my question if I do these three modifications on the home I want to buy, ⁓ does that delay the timing? So the more you can really get to the AI being able to act, I think it really transforms the sales funnel. Sandeep Ahuja: Mmm. Alex Wolkomir: in a lot of asset classes. Another very real example we're seeing is around facilities property management tasks. So things like when a maintenance ticket comes in, let's say there's a flood in the apartment, AI actually look at the ticket, have some judgment on what needs to happen, and you build a little agentic system where you might have one agent sole job is to figure out what the next best action is for a ticket. If that next best action is let's the particular stakeholder, there's an agent that picks it up and sends the email. If next best action is find a vendor, it pings vendor management system you have in place, looks at vendor ratings, at potential bids and costs, and is able to find someone to triage it. It an update to the tenant that we recognize there's an issue, we're having in place, and we're to care it quickly. But now you really start to automate that to a point where you might not need someone to actually touch the ticket versus it can be auto triaged. I think we're seeing similar things in construction world where when you think about everything that happens there from a project perspective, have the whole journey from the RFI to the submittal to approvals to schedule changes. And we're to think about all the things that happen in that project loop. it's all really around how do you not drop the thread and how do you keep things connected, you Sandeep Ahuja: Mmm. Alex Wolkomir: ⁓ draft documents at different points, update logs with what happened. those micro tasks are all things that AI agents are primed to be able to take on and sure those threads aren't dropped. And think broadly, I mean, all those examples stand deep in my mind, I think to your earlier point about use cases point to a reality of change is not happening at the small use case level. It's happening the workflow level it's how do you chain together Sandeep Ahuja: Yeah. Alex Wolkomir: use cases, chain together agents actually deliver the work in fundamentally new way. Sandeep Ahuja: Yeah, that's really fascinating the way you kind of bring the whole thread through, because I was curious. mean, wrote the playbook on use cases, and now you're like, no, no, no, no, no, it's not about use cases. It's about how use cases come together to a workflow. And the thing that I hear in many of these examples, at what I pull away, is also as consumers of, like you gave the example, if you're looking for a house or what have you. Alex Wolkomir: Hahaha Sandeep Ahuja: We're becoming impatient too, right? Because we do have everything on our fingertips. I can just order, know, like food and it better be here. I can order groceries and they better be here. can, gosh, I ordered some baby in the afternoon. They were here by the evening and I needed them to be here, which is my expectations have evolved. And I really think that that's that's the change that people Alex Wolkomir: Absolutely. Sandeep Ahuja: are expecting to see and people are observing, I'll give you my very raw, non-real estate example, non-architecture example. My dishwasher broke yesterday and I put a quote on this app Thumbtack where there's a whole bunch of different vendors can quote and then they'll come out and fix it. I live in this old 1920s home and I just needed someone who understood. that in the 1920s, hope anything is possible. So be prepared. the vendor I ended up picking was A, the one that responded the fastest, B, he clearly had automation and AI built in because it was asking me not, and there's a difference between AI and automation, but I could tell ⁓ that there actually some tiny layer of AI layered into his response messaging because it was a touch smarter. Alex Wolkomir: You Sandeep Ahuja: I was able to get everything that I needed, including a quote in like five minutes. And I going to come this afternoon, hopefully not during this podcast. ⁓ But that that means to say is that real timiness that makes it all so, so, so valid. Like I can see that that's what people are connecting to even with the tenants. Alex Wolkomir: Ha ha ha. Absolutely, it's all about, I think things. ⁓ One is big consumer change. and user behavior change, whether that's an enterprise tech, whether that's in consumer, but how we search, we interact with technology as move to more of these conversational interfaces, the way think about UX UI is totally different. And I think both opportunities and also tremendous risks. Cause like if your are going to be interacting, for instance, with let's say a voice agent with a textual agent, there's some point where you start to get to a risk of if everything going to be flat, soulless, what makes one company different from another? And how do you actually start to think about designing UX differently those worlds? the analogy I always use is, a few years ago, if you look at a lot of the big retail brands, it was very in vogue to move to black and white logo with sans serif font. And all of a sudden you walk through the mall, you look around and the logos are indistinguishable. I think there is a risk with a lot of the user facing AI. that it does become a race to the bottom. I do think there is an art to this of how I automate, augment, transform things that may happen behind the scenes, the enablers, the process moments, but how do I also preserve ⁓ those moments that matter, that define my brand, define engagement, that are really human and build that connection. And I say that's when we talk about, I say use cases ⁓ kind of gone away and it's about the families. I think what we're really seeing is you have to take what I call like a whole domain or family of use cases. But do you actually redesign ⁓ that that experience, that piece of work, really break into its little component parts? And like, what those steps that are really automatable, the tedious someone might not want to do, where AI can take care of that? And then what are the moments of humanity? Where there's a thought, ⁓ judgment, an application of taste, how do you actually, one, AI tooling to make person even stronger in that moment, but two, how you not get rid of that humanity ⁓ in that moment that's where it matters. you almost start to about the product is the experience, the product is the workflow. The product is not some AI model, it's the operating model that AI is a part of, but the answer to. And that's, think, the danger with the use case mindset is a lot of orgs are like I'm gonna do shiny little use case, give everybody a tool, but all does is add 20 minutes of work. It doesn't change the outcome. And they track it by saying, everybody logged into it today. And that's not impact. That's just, you know, performative technology in my mind, Sandeep. ⁓ Sandeep Ahuja: Yeah, a hundred percent. that resonates so closely. It's almost like you're describing our business model. Cause that's basically what we did to architecture, right? Where instead of building a point solution for some feature within architecture that architects could use or developers could use, we created the pipeline, like the steps and the order in which the steps go and which of the steps need human intervention and which of the steps don't need human intervention. across these hundreds of steps that, for an example, goes into designing a multifamily building from start to permits. that's process. That's our moat and IP now. It's not necessarily this tool and that tool. mean, sure, tooling is involved, it's not about tool. It's about the workflow and the order. And can you actually get that output? So when you say that, I'm like, Yep, makes perfect sense. So I would love to hear about you're seeing that happen more. Like where are you seeing the companies using the in a little bit more robust ways? And obviously no needs for companies necessary, but which workflows seem the most ahead from your vantage point? Alex Wolkomir: Yeah, I'd say... Some of the examples I gave earlier I think fall in that, but I think a lot around maintenance and facilities. everything when you think about ⁓ signal of something is wrong, the triage, access to the systems, dispatching folks come out to your home like you're waiting for today. And then all of the updates, the approvals, how you actually close it out, check the invoice against the accounts payable. But that whole I think is very prime. And I think we're seeing a lot of folks, whether it's everything Sandeep Ahuja: Mmm. Alex Wolkomir: from management in an office or an industrial warehouse through to in multifamily rental building. I think lot of promise and a lot of really exciting applications happening there. I'd secondly, going back to the leasing and sales piece I mentioned earlier, how do you start to you know, if I take the leasing example, really automate the lead to the lease journey end to end, including can AI schedule the tours? Can do the qualification? Can you help people match the buildings and really cement those moments as around it. I think in particular, a lot of the leasing those workflows have also proliferated quite bit because in real estate, a lot of times the tooling for things lives in a CRM and not in the core ERPs and systems record. ⁓ So the integrations are easier. And I think also a reality we've been seeing is given the way value from this AI flows through the NOI of the properties, has been a bias towards AI that drives revenue over AI that drives cost savings. ⁓ until AI becomes more ubiquitous, folks not going to underwrite in their acquisition models that assume operating models of the property if the new buyer doesn't have the same capabilities as the current owner versus folks will underwrite the higher rent roll all day. Sandeep Ahuja: Mmm. Alex Wolkomir: I do think there's been a bias in real estate towards, I'll the revenue focus versus the cost use cases like the owner operator end of things. I think it's different equation when we're talking about, you know, the facilities the pure third party operator, where they're thinking about what all the costs that sit on my books and how do I get to that better economic model. I think we've quite a bit on asset management as well, in terms of folks thinking about all kind of the manual tasks that happen around investment? how do you really bolster judgment moment of ICs and of investors? But thinking about for any investment I'm gonna do, the facts kind support that investment live everywhere, in different systems, in leases, different dates and key items are locked around in documents and all the different performance narratives. But how do you take away all work of having to hunt all that down and do it? Sandeep Ahuja: Mmm. ⁓ Yeah. Alex Wolkomir: spoke every deal and process. I think we're a lot of opportunity but I think it's all when done ⁓ right. It's how is it in way where the tech is it easier, the tech taking away all that tedious work, but it's really thinking about AI is not the AI is the enabler and also, the product is that workflow that includes the but also includes the people and the exceptions they're making it's the big brand choices or big money moves. How is the human in the loop with good information for really shifting ⁓ product mindset ⁓ that broader of how actually architecting a new way of say in the most extreme cases, I'm also seeing folks even rethink who they're hiring because of that. Because when you redo that workflow, if think about it, the ⁓ tasks someone doing of space today ⁓ look different when 60 % of that work is done through AI automation. And what the % that's left where it is thought, it's judgment, it's taste, it's relationships, I may recruit meaningfully in terms of who was profile that was really going to shine do that So I do think that's the other benefit that like domain and workflow posture. Is it an ideal world? I'm reinventing the world. how it gets done, bringing AI in to really elevate the experience, deliver it better. But also may think about the human moments differently, even in terms of how you're hiring to really the value. Sandeep Ahuja: That's yeah, there's so many different points that you touched on. ⁓ Everything from the front end on the sales revenue generation motion, where folks are leaning in first, which sounds kind of like obvious, like it makes sense that why people would start there, But it does sound like people are also investing in that whole Alex Wolkomir: Hahaha Sandeep Ahuja: ⁓ a tenant maintenance standpoint. And I like about the NOI you touched on it. That makes me think of, obviously in your research, you found that leading real estate companies have automated more than 70 % of tenant interactions, like you said, by using AI and they're seeing two to 4 % NOI increases. Now, obviously that's real money, but the number that really jumped out to me was that a 15 % Alex Wolkomir: Yeah. Yes. Sandeep Ahuja: rent premium between the best and the worst performers in the market driven by customers experience. how much of that is AI enabled versus just ⁓ good operations? Alex Wolkomir: Yeah, and think you're talking about the article we wrote, I think, two years ago on CX and customer experience. I'd say two parts to the answer. I'd say one is we think a lot about just the real growth of brand in this space. I think for any of the more consumer asset classes, know, multifamily branded residential, which we've seen booming in of like the number of hotels that you can now live in ⁓ quote unquote. think there's big part of it, which is the experience. I think where it AI enabled is a lot of what has now made that possible is AI lets you deliver Sandeep Ahuja: Yeah. Alex Wolkomir: those moments that might have been human intensive and by ⁓ really expensive as a tenant, but do it in a scalable way where I need the equivalent of hotel staffing to deliver hotel like hospitality. I've seen really players. How do you understand resident satisfaction, but not just take it at a face value versus when we have interactions and tickets, you run NLP on those and understand sentiments? Can you arm your front line? with simple things like recognize birthday or if you're going to give a renewal offer, you you might want to give them some type of service concession credit, but you want to give ⁓ someone with dog dog walking ⁓ versus someone has a parking space parking credit ⁓ that's an amenity on the property. But that personalization, I matters a lot that AI can power quite a bit. I think also like simple that can do and Sandeep Ahuja: Mmm. Alex Wolkomir: in terms of how does the resident feel heard responded to. I don't like the emblem is like, how you build the Domino's pizza tracker of X? And that could be for like a maintenance request, like, you know when your pizza has the cheese going on you know, when it's hitting the oven, you know, when it goes in the car, but that should be the same for anything. Like it shouldn't be, shoot a maintenance ticket because my lights out and it sits in the ether for two weeks. want to know what was seen. I want to know it's... ⁓ Sandeep Ahuja: haha Yeah. You're like, someone seen this? My light is still out. Come fix it. Alex Wolkomir: Yeah! And maybe like at the light bulbs out of stock, tell me that. And we ordered it and it's kind of come in and we'll install it when it's there. But lot of times it's communication and those moments that ⁓ where transparency adds a lot of value. And I think AI can be a big driver those moments. I think it drives it in two ways. is, the operational end, really letting you get to excellence a way that feasible, but it was on the experiential end of how do you things at scale. it may not have been possible. How do you use technology for those differentiated moments to kind of level up your team? And it may not be the consumer interacting with AI. It may be AI coaching your staff so everybody can level up and have that context. I think it's a big driver and I think will continue to be so. And I spent a lot of time in all different parts of real estate. So I also do a lot in hospitality where I think the same conversation's happening Sandeep Ahuja: Sure. Alex Wolkomir: hotels. How they get all of the great value but not lose the brand identity in those ⁓ moments matter with their And how do they also, know, in a world where for travel may be totally different, just like search for a home or search for apartment may totally change if it a conversation. What is their role? What is their brand voice? And how do they engage with the consumer in a different way? I mean, The big example might be I imagine a lot of these conversational consumer facing UIs, it almost creates a moment similar to OTAs and travel or the listing platforms and residential where it now a new ⁓ surface of engagement. there's a little of a question of if I'm the end provider of the service, what is my strategy around that new engagement surface? ⁓ Sandeep Ahuja: Mm. Yes, yes. All right, we'll bring it home with kind of the last and final personal rapid fire. So definitely don't think, just answer. Coffee or tea or is just life juice that you're powered on? Alex Wolkomir: say mornings a cup of coffee, but during the day tea. ⁓ a big herbal tea. Like I love a good peppermint tea and a ⁓ sprint mint. Yeah. ⁓ Sandeep Ahuja: Okay. Good peppermint tea. Morning person or night owl? Alex Wolkomir: both. I'd say a lack of sleep person. ⁓ My whole life I am wired where like it's very hard for me to stay asleep for more than maybe six hours no matter. Yeah. it's just ⁓ I was a baby like it was the bane of my parents existence going back way back when. Sandeep Ahuja: goodness of the above. And so now you're just using your powers for the good and solving all the AI problems in real estate. Alex Wolkomir: Yeah, Yeah, the answer is not morning or night, hyperperson is probably the way to think about it. Sandeep Ahuja: All right. spreadsheets or slide decks. Alex Wolkomir: ⁓ I'm a visual side person. Sandeep Ahuja: What is the most overused buzzword in real estate AI right now? Alex Wolkomir: Oof. I mean, still think it's probably, frankly, use cases. feel like, and like just over fixated. heard a great quote the other day that like too many real estate companies treat use cases like Pokemon where like it's you got to see them all, collect them all and like ⁓ successes. I've 150 of them, even if none of them have created ⁓ value. Sandeep Ahuja: Ha Love it. Use cases. think you're right. And lastly, what's the one AI tool that you use the most in a day? Alex Wolkomir: Personally, I do a lot of creative writing and poetry and stuff. I've been playing a lot more with like, Suno AI and some of the music models of how you actually translate play with it, and you can pretty sophisticated prompts around all the music properties. But it's really interesting to see the application of it. ⁓ Sandeep Ahuja: ⁓ yeah, I know. for Christmas, this totally random story. My husband made a song, he's from Indiana, so he made a song about driving home for Christmas. It was such a catchy song, it was 100 % AI. used, I wanna say sono or sono for the whole process, yes, sono. He used it for the whole process and now my little three-year-old likes to hum that catchy song that daddy made. Alex Wolkomir: You Tuna, yeah, yeah, yeah. I I'm gonna say I went to a conference recently where they made a real estate song with Suno to open up the conference and had all of the buzzwords you wanted me to call out earlier for sure, Sandy. Sandeep Ahuja: that's too funny. All so last thing, obviously you advise the biggest real estate firms on the planet. You see who's getting this right, you see who's kind of fumbling through it or taking the right steps, but not there yet. And obviously you're talking to real estate professionals every day who are trying to figure out how to navigate this moment. Some are excited, some are overwhelmed, and most are somewhere in between. What's the one piece of advice? you would give to someone in real estate who wants to get AI right, not just adopt it, but actually get value from it. Alex Wolkomir: I would say pilot 10 use cases, ship new workflow or a new domain. pick a thread of a workflow that has real volume and real pain for your people, it down to those steps and those thoughts that we talked about for different steps, build little atomic agents, automated aggressively, but also the moments of judgment, taste, humanity, protect them deliberately. I'd say get pilot to actual migrate into your systems, really have an organizational view of what are permissions, are approvals of when AI can act versus a human has to be in the loop. And also think about the change management. what I by that is it's not just we're to train people to act differently, it's how do you actually the AI tool and product with the change management built in. we've done a lot of implementations the step change thing is actually we're going build an extra screen where for two months, you know, the maintenance team going to hit approve. Sandeep Ahuja: Mmm. Alex Wolkomir: actually gets them to use the AgTool is you can go back to them and say, you hit approve 99 % of the time. don't we automate this? But there's little things where it's actually the product design that become unlock. Sandeep Ahuja: really cool. And the thing that really hits home for me is the value of actual intellect actual human thought and protecting that. I think that doesn't get talked about enough. And I think that's such an important point. Well, Alex, here. Partner at McKinsey & Company. Thank you for bringing the Real Talk today. This was so much fun. Alex Wolkomir: Thank you for having me, Sandeep. had a great time. Appreciate you having me.