Latest / State of the AI Union / Elon Salfati: Building for Adaptive AI Organizations
Transcript
- 0:06Hello everybody, and welcome to this week's episode of the State
- 0:09of the AI Union. I'm your host, Laura, and I'm
- 0:11joined with Elon Salfalti, who is the CEO and Co founder of
- 0:15Salfalti Agency. Really glad to have him here.
- 0:18And Elon, why don't you introduce yourself and tell us a
- 0:21little bit, a little bit about your company?
- 0:23Hey there. Hello and thank you for having
- 0:25me here today. So my name is Erin.
- 0:28I'm based out of Switzerland in in Zurich and I've been in the
- 0:32startup ecosystem for for a while now.
- 0:34I think about the past 14 years I LED in my life, you are in the
- 0:39departments, founded a few companies, had even a small
- 0:42acquisition in my life. And really so you know, all
- 0:47these years we have learned a lot around the intersection of
- 0:52AI, machine learning, cyber security and enterprise software
- 0:56and the problems that come in these spaces in particular,
- 0:59which led us to founding a company that we have today so
- 1:03far to go up. And the idea here is that we
- 1:06here constantly trying to challenge how AI adoption comes
- 1:11into play within large organizations and how business
- 1:17model are shifted with everything that we see right now
- 1:20in AI. So we're working on daily basis
- 1:23with all the way from like YC startups to Fortune 500
- 1:27organizations and we're trying to help them define what does a
- 1:33really an, an AI solution that stands out in today's market
- 1:38look like. That's awesome.
- 1:41OK. So one of the things that you
- 1:43said in your introduction was that you are helping customers
- 1:47all the way from Wyoming customers to enterprise Fortune
- 1:51500 customers. And, and in our prep we talked
- 1:56about how the like the business vision has like the time for
- 2:04business vision has shrunk very rapidly in in the last five
- 2:07years because of these LLMS and a lot of these frontier models
- 2:11that are changing. So and so how do you
- 2:13fundamentally, how do you think about fundamentally building a
- 2:16durable and adaptive organization when everything is
- 2:20just moving so fast? Yeah, it's, it's a great
- 2:23question as as you, as you said, it's moving so, so fast.
- 2:28And we can see it even historically, like if you if we
- 2:30look a bit back, you could easily see, you know, AP firm
- 2:36incorporating themselves, you know, 25 years ago and they
- 2:40could easily plot 20 years how it would look like.
- 2:44And today, if we just talk around really the last two weeks
- 2:49with an Tropic releasing 4.6 and open AI releasing what they call
- 2:55I think open AI frontier, they are releasing these small things
- 3:01quote UN quote and they are completely changing the market
- 3:05right. The the day after you could see
- 3:07so many companies are dying and we can see it even in the stock
- 3:09market, you know, organizations like Figma practically dying
- 3:14these days. So how do you really like the
- 3:16big question is how do you build something that is doable and
- 3:19adaptive? And we like to see ourselves as
- 3:22an AI native agency. And when I say AI native, I
- 3:26really mean not just that we have a lot of AI tools
- 3:29internally. That's that's nice to have, but
- 3:32it's not really the core value really.
- 3:35What I mean by that is we have an infrastructure internally
- 3:40that is designed to change every three months.
- 3:44Like this is this is the mindset that we are really building to
- 3:47trying to build, right. What does go to market look like
- 3:51if you change it every three months?
- 3:52What do you get known for if you change the business every three
- 3:55months? And how do you kind of
- 3:57constantly adapt and build something that is of value?
- 3:59The way that we are personally looking at it is if we look for
- 4:05a second on what LLMS are, they are really an echo of humanity.
- 4:12And the one thing that they are lacking, you know, for a really
- 4:17like deep mathematical perspective is the ability to
- 4:21create. They are not creative.
- 4:24So everything that we do in the organization is trying to
- 4:29understand what is a commodity that can be delegated really to
- 4:34a quote UN quote army of agents and what is it the things that
- 4:37we really have to to focus on and advance.
- 4:41And for us, it's the research, it's the it's getting the
- 4:46clients that work with us to really focus on outcomes, to
- 4:50really focus on a strong added value that is constantly ahead
- 4:54of the curve. We want to move fast as fast as
- 4:57possible. Yeah, I like what he said about
- 5:01you know the the infrastructure designing an infrastructure that
- 5:04is designed to change. Can you tell me more about what
- 5:06that looks like just as you said in like go to market, what is
- 5:10what is it? What is go to market
- 5:11infrastructure looks like that is designed to change like every
- 5:14three months. That's, that's a good question.
- 5:18And, and we, it's, it's also a question that I don't think we
- 5:23will ever be able to completely figure out like we're struggling
- 5:27every day. But I think, I think this is
- 5:29part of the, the constant right if for it to change every three
- 5:33months. We need to really to struggle
- 5:34every day. And the way that we go about it
- 5:37right now is that we are trying to make sure that we get known
- 5:42for something that is timeless, right?
- 5:46And for us, our anchor value is what we call AI, native human
- 5:521st. And you would notice that we,
- 5:55funny enough, put the human first, literally 2nd.
- 5:58And the idea here is that we would want our organization to
- 6:03be led by AI as much as possible.
- 6:06And for us, that also means that we are no longer subscribed to
- 6:11HubSpot. We are no longer subscribed to
- 6:14any of the common tools that we see out there in the market
- 6:16because they don't fit our use case as much as we need, right?
- 6:21For us to have an edge, for us to adapt every three months, we
- 6:24need an environment that is an echo of us.
- 6:31I cannot change my processes to fit HubSpot or Salesforce or
- 6:36whatever, just to you know, quote UN quote convenience.
- 6:40It's holding us back right now. Yeah, like I like what you said,
- 6:44in an AI native company, you have to find vendors and
- 6:47solutions that are going to fit what you are doing, right.
- 6:52Let's let's flip that around a little bit.
- 6:53You know if the go to market infrastructure also has to
- 6:55change, we are also seeing right now that buyer infrastructure or
- 7:00buying decisions are changing somewhat.
- 7:02The era of your experimentation is over a little bit because
- 7:06experiments have not done really well, right?
- 7:10Something like 90 plus percent of AI implementations failed.
- 7:13And so when when customers or buyers are running a cycle
- 7:17today, what are some specific decision criterias?
- 7:21CR OS should demand or CE OS should demand.
- 7:25CF OS should demand to make sure that we're just not
- 7:29experimenting with another, you know, chatbot.
- 7:33Yeah, the the infamous prototype and exploration purgatory, it's
- 7:39like you said it and I think you put it nicely, just trying
- 7:42another checkpoint is really not, not what we are trying to
- 7:45accomplish. So the way that we have been
- 7:47looking in that and I actually have a nice example is we mostly
- 7:54look on AI solutions as a way to, I'll call it boost what we
- 8:01have today. This is how most organizations
- 8:04go into the process and they say, you know, I have like this
- 8:07plus especially in enterprises, I have like this process that
- 8:12spends across half of the company.
- 8:14It can easily take me 6 months to accomplish.
- 8:18I will try to have like this one piece, one step of the
- 8:23milestone, I'll try to optimize it.
- 8:25And they usually default to looking at, at the parts of the
- 8:32process that are the most convenient or familiar to them.
- 8:36So the most, in other words, deterministic.
- 8:40But really this is not where AI efforts and exploration except.
- 8:45So we have worked with partnership with Tribe AI, it's
- 8:48another AI agency. And we have delivered for Key
- 8:54Loop, a solution. Key Loop is a UK organization.
- 8:57And as an organization, what they do, they're connecting,
- 9:00they're basically the data connectivity layer between the
- 9:04car manufacturers and the dealerships.
- 9:08And they have this really long process on how they would take a
- 9:12request from the car manufacturer and pass it on to
- 9:16into the system and then to the dealership.
- 9:19We have sat down with them for a workshop for two weeks and the
- 9:24idea was very, very simple. We are starting with a clear
- 9:28anchor at the company level. So we chose as a metric for the
- 9:32success of the prototype the churn reduction.
- 9:36And so we worked in collaboration with the head of
- 9:39churn. And what we have seen that was
- 9:41very, very successful though is we are not trying to take a
- 9:48process that is, you know, 80% or 90% optimize and make it 100
- 9:55and take the human out of the process.
- 9:58What we really want to do is we want to look at the processes
- 10:01that are heavily human driven, very inconvenient for everyone
- 10:06to do. And usually if you look at the
- 10:09company, humans would do them at like 10 percent, 20% success
- 10:15rate. It's usually the processes that
- 10:17really no one likes to do but have to do putting invoices into
- 10:22the CRM and then filling all the fields in whatever system that
- 10:27they have. These things tend to have a lot
- 10:29of mistakes in them and almost 0 value for the person who inserts
- 10:33them into the system. When we started this pilot with
- 10:36Kilo, I will expected target success rate was to take the
- 10:43process from it's OK 35% of the time, which is what they had
- 10:48before we started and take it all the way to 65.
- 10:52At 65 it already had massive value for them, but the nice
- 10:57part about AI is that we automatically took it to 85%.
- 11:01Wow. It's much harder to fail when
- 11:05these are the games that you are playing on, these are the AI
- 11:08pilots that you really want to tackle.
- 11:11So what I'm hearing is, is that we should be building right for,
- 11:16for new processes by rotating over rotating on the business
- 11:20outcome versus just changing the, the way versus just adding
- 11:26AI onto on top of the way that you're doing things today,
- 11:29right. And you know, one thing that we
- 11:31had previously discussed before is the way to do that.
- 11:35And, and you, you said as well that, you know, a lot of legacy
- 11:38software doesn't fit current use cases.
- 11:41And one of the ways that we can do that is we can actually build
- 11:44custom software for companies because now with, by coding, you
- 11:47can up a software that you want. You don't need to ask the
- 11:51company to give up their DNA, the way that they do things, the
- 11:54operational excellence that they have, you know, built to
- 11:57actually sustain their company. But just like, just like in this
- 12:03example you gave us where we were asking companies to do
- 12:08things differently, right? We're actually the process that
- 12:10is human oriented and, and asking them to kind of change it
- 12:13and make it AI optimized, machine optimized.
- 12:16I always talk about, on this podcast especially, we talk a
- 12:19lot about how AI moves at the speed of humans and it's only
- 12:22successful when humans are willing to give up their control
- 12:28or the way that they're doing things.
- 12:33How, how do you reconcile that? How do you reconcile building
- 12:36software trained to deliver an organization's business needs
- 12:40without giving up their execution power, which is the
- 12:44reason why they're successful in the first place?
- 12:47Yes, that's it's, you know it's, it's a fun when I kind of put
- 12:52out myself for alcohol. I thought about APE firm that we
- 12:57are working with. They have been in the market I
- 13:00think 25 years. They are really the classic
- 13:04definition of legacy software. They have built everything
- 13:08internally. They almost haven't bought
- 13:11anything, but they stayed true to themselves to the last like
- 13:16smallest Unity can you can think of.
- 13:19And funny enough, they are the best organization or the most
- 13:24well positioned organization to jump into this AI native ship
- 13:30because they are already so true to themselves.
- 13:32And if you think about it, actually don't think that these
- 13:35two two conflict like ownership of what humans want and having
- 13:41like an AI native organization. It's not about taking away the
- 13:47humans, right? Again, like if I go back to what
- 13:50we started our conversation around like this AI, native
- 13:53human 1st and human creativity, this is really about delegating
- 13:59to the individuals the ability to think broadly, to think
- 14:05creatively. I cannot find a single employee
- 14:09in a, any of the organizations that we work with who say, you
- 14:13know what? I love to fill in forms like
- 14:16This is why I'm here for. This is why I wake up in the
- 14:18morning. It doesn't happen and you know,
- 14:21these things change, right? Things that used to be really
- 14:25challenging are now becoming a commodity.
- 14:29AI for me is here to allow us to think broader over over dealing
- 14:38with a process. So if we look, I always tell to
- 14:43to all my employees, like the way that I'm thinking about it
- 14:46is, you know, if you look, I don't know, 100 years ago you
- 14:50had the Library of Alexandria. And you know, scholars would
- 14:54spend years and years of really lifetime going over the library
- 14:59trying to deduct maybe one insight and very likely that
- 15:04someone in a gigantic library on the other side of the world did
- 15:07the same thing. Google came in and the Internet
- 15:10came in and now you could spend years finding something online
- 15:16and deducting insights, but it's not a lifetime.
- 15:19It's like much shorter. Now with AI, we really have
- 15:24human knowledge at the people of our hands, and it really just
- 15:29means that adapting and really the muscle memory of learning is
- 15:35so much more important than the actual thing that you do.
- 15:39I see a lot of articles right now about big organization like
- 15:43Meta replacing management layers with AR make sense on one side.
- 15:49But on the flip side of it, really often like the
- 15:53management, the senior people are the people who are able to
- 15:57looked very broadly on the organization and very abstractly
- 16:02on how the organization can operate.
- 16:06And these are the things that I really don't want to lose on.
- 16:09Like the muscle memory of learning and adopting is really
- 16:13the key, and fitting everything to how we work is what makes
- 16:18this change so unique. So, you know, that's really
- 16:22interesting, right? Because what we're saying is
- 16:24we're saying, OK, we now want to be looking at employees as like
- 16:29human capital should now be focused on learning creativity,
- 16:35being able to embrace or use AI in a way that's hasn't been done
- 16:39before because only humans can create to your point, right?
- 16:44And Salesforce as, as an example, created a whole
- 16:50ecosystem of employment around, for example, their
- 16:53certifications and these skills that that people have spent like
- 16:58years mastering and learning. And they're worth quite a lot of
- 17:03money. You know, these are knowledge
- 17:04workers, right? And now that knowledge is at the
- 17:07tip of our fingers. What what do these
- 17:10certifications mean then in terms of transferable skills?
- 17:14When you move from company to company, does it mean that, you
- 17:16know, you should no longer put those certifications up?
- 17:19How do you do you demonstrate? How demonstrate?
- 17:22OK, I have learning. I have learning and creativity
- 17:25in my DNA to a company that's trying to hire you when sort of
- 17:30everything that we've done before is I have certificates, I
- 17:32have proof of the knowledge, you know, of the things that I've
- 17:36done. Yes, it's a fascinating question
- 17:40And I would even dive a bit deeper and I would say like if
- 17:44you think about it right now, like the next AI native
- 17:50generation is born, like what do their life look like?
- 17:54And really it would not be about the knowledge, right?
- 17:58It would be around like the discovery of knowledge.
- 18:00So yes, I don't think I assume many people won't like to hear
- 18:04it, but I, I assume that, you know, another sales force
- 18:08certificate is not what's going to be the the edge anymore.
- 18:12Judgement is, but I don't think that these certificates take
- 18:17away from demonstrating judgement, but judgement is the
- 18:21thing that you would want to demonstrate.
- 18:23I would value someone who have certificate from sales force,
- 18:27HubSpot and ATIO over a person who has like 5 certificates from
- 18:32Salesforce. OK, so does that mean we are, I
- 18:38don't know in, in a race to the bottom or you know, judgement is
- 18:43becoming so hard or like how to measure creativity?
- 18:48We you know, we have done that before in other spaces, right?
- 18:52So if you look on research jobs or if you look on design jobs,
- 18:59there are ways to manifest creativity.
- 19:02Management jobs are ways to manifest creativity.
- 19:06Tell me what's 1 like interview question your favorite interview
- 19:10question to sort of discover this in a higher.
- 19:15That's a good question. It's not a favorite question,
- 19:19more as a trying to, not coming from a bad place, but like
- 19:25trying to tackle thinking. So for example, when I'm
- 19:27interviewing engineers, I would often ask them questions in a
- 19:33programming language that they don't know.
- 19:35Because for me, when I look at software development, which is
- 19:39honestly these days a commodity, it's meaningless.
- 19:42But yeah, give give me like 2-2 years.
- 19:46And so two years ago, I would ask them questions around the
- 19:49software language that they don't know.
- 19:51And the idea here is that when I think about software for me, you
- 19:56know, you have iterations of the code and you have data that is
- 20:00coming in and data that is coming out, and you can bundle
- 20:04them into blocks that are called functions.
- 20:07It's so abstract for me. If someone would ask me to write
- 20:11in this language over this language, it's meaningless.
- 20:13I want to see that thinking in the people that I hire.
- 20:18So I take the same concept into. Really every department that I
- 20:23hire for. So if I go to marketing, I'm
- 20:26trying to think and questions and ask these questions around
- 20:32areas that they are not familiar with.
- 20:33I would want to see how they deduct knowledge over
- 20:40demonstrate knowledge. How to deduct knowledge versus
- 20:44demonstrate knowledge? OK, that's awesome.
- 20:45I really like that. OK, last question before we move
- 20:48on to something slightly different.
- 20:50So one thing that you also said earlier in this conversation
- 20:54during our prep was we will likely or hopefully we'll move
- 20:59to a model where instead of a human hiring an army of agents,
- 21:03we'll move to an army of agents hiring a human.
- 21:07Big question here. And I know we had a joke about,
- 21:10you know, just using the human for like the biometrics and the
- 21:12fingerprint. And, and I think there was a
- 21:16video about that also. But you know, as, as, as, as I
- 21:21think we give up more to the, to the LMS and to the, and to the
- 21:24AI and allow them to do more, right?
- 21:27What do you think this does to the human element of trust and
- 21:32accountability? And I see that both from a trust
- 21:35factor for, for like the consumer of the A, like do I
- 21:38trust, do I trust an army of agents, right.
- 21:42And then who's accountable for the army of agents, if they're
- 21:46the ones that are hiring the human, you know?
- 21:50It goes like the accountability question, as you put it, is such
- 21:55a great question. It goes to so deep to human
- 21:58psychology. You know, I, when I think about
- 22:01it, I again, we were saying that we are an AI native human first
- 22:06organization, and yet we are trying to delegate as much as we
- 22:10can. Today.
- 22:12I, I, I don't feel like I give up control.
- 22:16I feel like I gain it actually because now I get to steal the
- 22:21ship without dealing with, you know, the steering wheel.
- 22:24And if we look on on the same question in other areas, right.
- 22:29So airplanes were able to quote, UN quote, fly themselves for
- 22:34years, but we still have the pilot in it just because humans
- 22:38feel the concept of a plane without human so.
- 22:42On that question, right. On that answer about autopilot,
- 22:46right, Do you think that the passengers, the plane knew about
- 22:51autopilot? I almost think that that many
- 22:54people still today, when they fly a plane, they think that the
- 22:56pilot is flying the plane. That is, I hope they know.
- 23:04There's a pilot, you know, he greets him.
- 23:06He says, hey, welcome on board. And you know, he, there's a
- 23:08human voice to it, right? Is it the autopilot is hiring
- 23:11the human to just say these words?
- 23:13And I'm just wondering up front, like from the perspective of the
- 23:18of the consumer, like you know, say, say, say you, you as the
- 23:24operator have more control because you have are able to
- 23:28have a greater span of control, right.
- 23:30But as the consumer receiving that end product, how do you
- 23:35think it benefits them to know or not know how much AI is
- 23:38behind this? I think that the biggest
- 23:41manifestation of having a lot of AI around it is that it gets so
- 23:46much more time to human connection.
- 23:49The more you don't need, if you think about the marketing
- 23:53organization, the more you don't need to manage, you know, Google
- 23:57Ads, the more you get to go and meet the people around you.
- 24:01And I think it's actually, I had a good friend of mine,
- 24:05Angelique, she was the previous CMO of PepsiCo, brilliant mind
- 24:10around everything related to brand and marketing.
- 24:12And we're constantly having this conversation around Internet
- 24:18protest, globalization. Is AI bringing us localization
- 24:21back? Like is it that it would be so
- 24:24much easier for me to have business relationships with my
- 24:28neighbor because we are able to meet each other every day over a
- 24:33person on the other side of the world?
- 24:35Because this trust that we have in one another is really what
- 24:44what gives us this control. Like again, if we look, take the
- 24:46same example of the plane, it also applies for autonomous
- 24:51driving, right? You have like that one car crash
- 24:55and there is no one accountable and everyone is fearful.
- 24:58But in practice, there are way less crashes from autonomous
- 25:01cars. Yes, yes.
- 25:04So I think there are two items here is 1A, we are able to have
- 25:09more human connection which I think is actually in powering
- 25:13it's enabler. But then if we go back to the
- 25:16question about embedding our DNA into this army of agents and
- 25:22actually not necessarily taking a product that shapes your
- 25:26process, but you're shaping the products based on your process,
- 25:29then your organization become an echo view and it becomes much
- 25:34more connected to your customers.
- 25:37And by the beginning of the conversation, I've mentioned the
- 25:40muscle memory of learning is so important right now, but it not
- 25:45just for us humans, it's also for the organizations, right?
- 25:48The muscle memory of the organization learning and
- 25:51adapting and meeting the customers the way that works,
- 25:55which always will be human connection because people buy
- 25:59form trust. That's not going anywhere.
- 26:02If I'm hearing you correctly, I think what you're saying is that
- 26:06with, with, with the AI, it will allow us to form closer human
- 26:11connections, which is the thing that builds the trust.
- 26:14And since we are changing the way, you know, software works by
- 26:20making sure that we implant our DNA into the software, then it
- 26:24actually should as a result, create more trust.
- 26:27Because it's not just like stop, stop solutions.
- 26:32It's something that is actually a replica of us.
- 26:34And we have. We have more time to sort of
- 26:36explain ourselves and build our brand as humans to the other
- 26:40party. Well put.
- 26:42Awesome. Elon, let's pivot a little bit
- 26:45and learn a little bit more about you.
- 26:48I'm very curious. I actually never met a founder
- 26:50from Switzerland before. We have a lot of them here in
- 26:52Palo Alto, but tell us about what it's like in the startup
- 26:56ecosystem over there in Europe. I love it.
- 26:59I actually have a series of talks at the UK Parliament, the
- 27:02House of Lords about that and how we do like AI adoption in in
- 27:09Europe. And especially now with some
- 27:12progress in the EU Commission and the 28th regime, I think it
- 27:17would open a lot of opportunity for investments into the into
- 27:21the US market. But I think you know, from a
- 27:23founder perspective, there is a really big benefit because we
- 27:28are very close, especially for our model like this AI native
- 27:31agency where we try to be exactly in between SAS and
- 27:35custom software. So you kind of get the benefit
- 27:38of both sides, but none of the limitations.
- 27:40Organizations here in in Europe have been around for years and
- 27:44years and years. They are the definition of
- 27:48organizations that have a lot of opportunity to change and scale.
- 27:52But on the other side, they are to your point about like how we
- 27:57build trust. They are the people that you,
- 28:00you get to do business if you're Dell with them having dinner
- 28:04three times a week, talking no business at all.
- 28:09So I think it's, it's very different in, you know, from the
- 28:13from the US market. But you know, we are neighbors
- 28:17and everything is very close and the collaboration is very
- 28:20important. And you can see a lot of like
- 28:22big player coming in. So that's I'm very happy about
- 28:26it. It, it is actually a, a very
- 28:28good reason region to like practice trust because that,
- 28:31that culture of, you know, and just like you said, like these
- 28:34organizations have been around for a really long time, a lot
- 28:37longer than I think like some American organizations, right,
- 28:41in general. And, and, you know, exercising
- 28:46that trust muscle, I think there's no, no better place to
- 28:48do it. You know there's a famous saying
- 28:52where America innovates in, China imitates and Europe
- 28:57regulates, right? Unfortunately.
- 29:00What what, what are you seeing as as some good things about
- 29:05sort of like that saying and some things that are challenging
- 29:07for you as a startup founder? Yes.
- 29:10So it is much higher bureaucracy in how we operate.
- 29:16So that that's a challenge that we see every day.
- 29:20I used to have like AUS entity and you know, when I started
- 29:24that company, you know, I go to Skype and click a button and you
- 29:28have a company here, you know, you need to go by hand and put a
- 29:34physically signed document at the notary.
- 29:37It's a long process, but on the flip side, I think it's really
- 29:43designed around making sure that individuals are empowered.
- 29:48So what I mean by that is privacy regulations are hard and
- 29:52the EUAI Act is hard, but I think it makes sure that every
- 29:58organization is is in check and grow in the right direction.
- 30:03So for me, those are here to well, it can definitely be
- 30:08optimized. Don't get me wrong, they can
- 30:10definitely be seen, but they are I think designed to help us grow
- 30:15in a direction that kind of benefits humanity overall.
- 30:19And you do see like a lot of brilliant founders coming in
- 30:23from from Europe. What I do hope EU and the
- 30:27European Union general will do is is give them the ecosystem to
- 30:32also scale you. So we have a lot of companies
- 30:35that start here, but the vast majority of them eventually end
- 30:39up moving to the US to scale, which, you know, on the long
- 30:43term, I think it's it's good that you're also as a partner to
- 30:47the US, right? Will have like the legs that
- 30:50they need to stand on. OK, So very interesting.
- 30:55Elon, thank you so much for your for being here on the show.
- 31:01You know, there are lots of topics we talked about trust
- 31:04and, and really interesting to hear perspective on being in in
- 31:09Europe and, and I love to see you know what, what else you
- 31:14guys build and for customers And do you have any last words
- 31:18before we close out the show here?
- 31:21I would just go back to our values.
- 31:23I would always go back to, you know, this AI native human first
- 31:28concept for us, like being adaptive and making sure that
- 31:32everyone who listens, you just stay focused on the outcome.
- 31:35Making sure that it's not just experiments and that you feel
- 31:39empowered and not skilled is what we're trying to do.
- 31:42Awesome Elon, thank you so much. And to all our listeners, if you
- 31:45enjoyed this episode, send us a like, drop us a comment,
- 31:48subscribe so we know to send you more.
- 31:51Otherwise, we'll see you next time.
- 31:53Bye. Thank you.
- 31:54Thank. You.