Omer (00:16.640)
Welcome to another episode of the SaaS Podcast.
I'm your host, Omer Khan, and this is the show where I interview proven founders and industry experts who share their stories, strategies, and insights to help you build, launch, and grow your SaaS business.
In this episode, I talked to Dennis Mortensen, the founder and CEO of X AI, an artificial intelligence driven personal assistant that schedules meetings for you.
Dennis had just sold his company and now had plenty of time on his hands.
When we have more time than we know what to do with, we sometimes do silly things.
And that was no different for Dennis.
As he sat in front of his computer, he opened up his calendar and started counting the number of meetings he'd had in the last year.
And he counted a total of 10, 19 meetings.
But what really shocked him was that 65% of those meetings that he'd set up had to be rescheduled.
So that got him thinking about a solution.
But there were already lots of SaaS products that help people to schedule and reschedule meetings.
And he knew how easily he fell in love with new business ideas.
So instead of trying to validate his idea, he spent the next few months trying to convince himself not to work on it.
When he couldn't talk himself out of working on the idea, he decided to move ahead to the next step.
And he validated his idea by building a concierge mvp, which turned out to be really smart.
And we cover exactly what he did to test his idea without writing a single line of code.
And that's how X AI was born.
Today, the company has raised over $44 million.
Dennis is a serial entrepreneur who's already had three successful exits as well as one failed startup.
And he shares some fascinating insights in this interview.
I think you'll really enjoy it.
Dennis, welcome to the show.
Guest (02:12.960)
Thanks much for having me.
Omer (02:14.560)
So do you have a favorite quote that you can share with us?
Something that inspires you or gets you out of bed every day?
Guest (02:19.680)
I wish I could attach this to somebody super profound, but very early on, Diesel, the fashion brand, had some campaign where they were putting out sweatshirts and what have you with this?
Only the brave quote that stuck with me.
It's probably 20, 25 years ago, and somehow I bought the sweatshirt and the sweatshirt disappeared quickly.
The quote, though, stuck with me.
And whenever I think about anything entrepreneurial or anything which you and I do in this vertical, that particular quote comes back to me.
Only the brave.
Omer (03:03.140)
I love it.
I think I actually remember that as well.
We're showing our age here.
Guest (03:06.980)
Yeah.
Omer (03:07.380)
Sadly, now you are originally from Denmark.
You are a Dane in New York.
How long have you been in the us?
Guest (03:16.750)
So we sold our prior prior company to Yahoo in May of 2008 and I arrived in New York in July of 2008.
So that's about 11 years.
Omer (03:33.230)
And what was that company that you sold?
Guest (03:35.150)
We did enterprise web analytics.
So think omniture web trends, Google Analytics, if you will.
Cool.
Omer (03:43.100)
Yeah, I mean, I would love to talk about that too, but I want to make sure that we cover the core kind of the XAI conversation and then if we have time, I'd love to jump back to that.
So for people who aren't familiar, can you tell us about X AI?
What is it?
Who's it for?
What problem are you trying to solve?
Guest (04:03.500)
Sure.
So we brought the team back together first weeks of 2014 with this idea that we might be able to go create these two intelligent agents that could schedule meetings on your behalf.
So this whole idea that Susan sends me an email saying, hey Dennis, I'm going to be in Manhattan in July.
You got time for us to kind of meet up for a diet cook.
And as soon as I say that, you can immediately imagine what happens after that because you've done that ping pong yourself and a thousand times.
And it's not that it's difficult, it's just yet another chore that you are forced to do as a knowledge worker.
And really the only way you can escape it in this day and age is that you stay 20 years with the same company, become SVP of whatever and you get a human and then the pain doesn't even disappear.
You're just giving it to Tom and now it's his pain.
So the agents that we've been able to kind of craft are these machine agents for you can say, oh, you know what, I'd be happy to meet up.
I have CC in Amy and she can help put something on my calendar for half an hour on 200 Broadway.
When you're here, I then click Send.
And remember, the agent is just an email address and I cc it in.
And as I click send, the agent then understands, who is Dennis?
What did he just ask me to do?
As I understand that, let me remove him from the conversation.
Let me reach out to one or more participants.
Let's make sure I drive that conversation towards conclusion.
And upon conclusion, let me insert that event into his calendar and to all of the participants.
Participants for this particular meeting.
And that sounds like one of those tiny things for where that can't be all.
Dennis, why are you so many people working on this for so long when it's just such a little chore.
But it's kind of similar to, say, Dropbox.
You're just saving files in the cloud.
How hard can it be?
Surely if you want to do it right, it becomes kind of hard.
But that's what we've been working on for the last five years.
Omer (06:01.080)
So tell me how you came up with the idea for this business.
Guest (06:04.200)
So our prior company, which was a predictive analytics venture for media, got acquired in 2013.
And whenever you have your company acquired, you end up with a little bit of money in one pocket and a lot of time in the other pocket.
And when you have a lot of time, you start to do silly things like going into your calendar.
Can I see?
Hey, so how many meetings did I do the year prior?
I hired some people, I raised some capital I didn't exit, and that added up to a staggering 10, 19 meetings.
And you know what?
What was even sadder, 672 of those very meetings that I set up myself got rescheduled.
And I think that was certainly a catalyst to kind of look closer into whether, is this the future?
Because I've now been kind of in the workplace now over four ventures, doing this for 20 years straight.
Do I just do this for another 20 years and then I die?
That doesn't sound right.
And that's kind of where we start, is going to look into how could one potentially kind of do something that could see this pain be removed?
Omer (07:22.720)
Okay, there are a lot of companies out there who are doing sort of scheduling, kind of giving you a calendar, sort of a booking page and so on, but you decided to take much more of an AI driven approach to solving that.
What kind of validation did you do?
Like, how did you get to the point where you felt confident enough that you were going to go all in with this idea?
Guest (07:46.880)
So I think any idea, just as a footnote here, as an entrepreneur, is not one for where you should try to fall in love with it.
You should try your damnedest best to persuade yourself to not do it.
Because you and me will fall in love instantly if you're just a little bit entrepreneurial and you walk up to me tomorrow and say, hey, Dennis, I think we should do a bakery in Westchester.
And you gave me three good ideas for why that should be the case.
Or I fall in love.
Yeah, we should.
Yeah, that current option is kind of shitty.
Let's kind of go to the whiteboard.
So falling in love is the easy part.
Discarding an idea is a little bit harder.
So to kind of answer your question kind of with that footnote in place, the first thing I did was to set up an email, agree with my VP engineering that for the next good month or so, you and me are going to play each other's assistance.
I wanted to see whether that joy of having somebody else do this for me was as dramatic as I imagined.
Because I've always been too frugal to ever have anybody doing this for me.
I'd rather sit at 11pm in my underwear no matter how much funding I've raised and do my own meetings.
So I wanted to experience that joy.
So we just set up an email called Amy and then Alex would schedule my meetings and vice versa, I would schedule his.
And you know what?
That was joyful, but it was also slightly painful because not only would I get the joy of having Alex kind of do my meetings under the kind of Amy email, but I would have to do his.
And I fucking hated it.
As in, it's not even my meetings, but I'm sitting here scheduling yours.
And that was not fun, but it was certainly something where I could see there's some joy here.
It might just be me.
So let me take the next step.
Then I hired a full time assistant and called up 50ish friends and said, hey guys, I know you are all early in career or middle management and you don't have anybody doing this for you.
I have paid for one and she will do it for you and I would like you to do it within these constraints.
As in it's just an email, you can't phone her or anything like that.
And you must kind of act in this particular way.
Because I wanted to kind of see if they would experience the same joy and I wanted to see if there was any patterns that would potentially arrive.
Such as these people are just such unknown entities that she couldn't actually do the job or do the job well.
Because if she couldn't do that well, there was no chance in hell we could ever get a machine to do it.
So she did that for a couple of months and they loved it.
As in, not only am I not paying for it, Dennis is paying for it.
That's awesome.
Nice to see him spend some of that money.
2 I need to kind of see this to turn into a product.
And that was the second step.
And then we can go on and on.
But those were certainly the two early ones that are used as qualifiers for whether it was just one of those ideas you have in your head.
But it kind of validated it through very cheap ways.
That's less than 10k spent on trying to kind of see whether this was crazy.
Omer (11:04.680)
So you effectively created like a concierge or wizard of Oz type mvp.
And as you went through this process, you kind of talked about, like, not falling in love with your idea.
What do you say to yourself as you're going through this process to remind yourself not to do that and to constantly be trying to talk yourself out of working on that idea?
Guest (11:25.520)
So another footnote, certainly, and I'm sure many of those that listen to this will have similar kind of processes.
But again, kind of coming back to the point from before of where you and me as entrepreneurs will fall in love easily and we'll see things in the world that are not right all the time.
Like standing in line at Citibank thinking, this is shit.
I need to start a retail bank.
You probably shouldn't, but that's what you're thinking.
And instead of having those thought kind of just running around in my head, what I do almost immediately is that I take that idea and I write it onto a list which I don't look at, just to get it out of my head.
Now what happens then over, say, four years for where I'm running a venture?
And before that comes to an exit, I just keep writing things onto that list.
And then when it's time, I start to kind of look at that list.
And this whole idea of meetings being a pain was on that list multiple times.
So as I start to kind of do some of those initial thoughts and do some of that initial testing again, to kind of come back footnote over.
To kind of answer your question, I actually invite in a set of friends, not to pat me on the back, but to kind of say, I've done some initial research here, I've looked into it, I've tried to kind of tell myself why it either cannot be done or it shouldn't be done.
And I'm still not in a position where I can kill it.
Your job now is to kill it for me.
Because if you can come in with very little prep and kill it, then either I wasn't prepped enough or I actually did fall in love and I fooled myself.
So it's to be one way.
I'll call in a set of scientists, a set of engineers, a set of commercial people, depending on the type of project.
And their job now is to sit in a room over a number of days to see if they cannot kill it.
If it survives that session, then there Might just be some meat on the bone.
Omer (13:19.620)
Yeah, totally.
Hey, I'm still excited about the bakery we could start.
So I think hiring a personal assistant and going through and sort of testing the idea, I think was really smart.
But to actually build an AI driven personal assistant that actually works well is not a small feat, right?
I mean, that's quite a challenging thing to do.
So how long did it take you to get to a point where you could kind of get the first iteration of the product out in the hands of users?
Guest (14:02.410)
So I'm sorry that I'm all about footnotes.
That's actually the first time I've done a footnote type chat.
But I would like to add another footnote here for where again, I'm sure most of us on this chat here, those listening and you and I would agree that most startups will take one of two risks, and I know this is oversimplifying it, but they'll take a technical risk for where there's uncertainties about whether they can actually go engineer the very thing they want to engineer, or they'll take a market risk for where surely everybody can go engineer an Airbnb.
But we're not sure that the market is willing to adopt the thing that you engineered.
And you need to kind of know which one of these two risks you are taking.
And if you're taking both at the same time, then you're certainly in the only the brave category and perhaps a slightly insane category.
We are obviously in the or I would like to believe we're in the technical risk category for where if we can pull it off, it seems likely that the market will adopt it if we can find a fair price for it.
And in that technical risk bucket, when we kind of go look at that, our trajectory compared to the traditional startup or certainly the market risk startup is slightly different.
So just because we were able to have the idea survive doesn't suggest that we had proof that it could be done, given that language is not a solved science.
So what we did was we assembled the team and we actually raised our seed round on the premise that we will now go out for a manageable pool of money and collect a data set for we can prove that this can either not be done, we now know for a limited amount of money, or two, it can be done.
And we certainly have some idea of the magnitude of effort that needs to be applied to do it in full.
And that means our kind of initial kind of seed or even kind of angel and seed was actually just a milestone where the only Outcome was a thumbs up or a thumbs down.
No mvp, no customer, no revenue, nothing but a I can come back to you and say, I think we can do it.
And that is certainly something that is different.
So you see what many of us have done in the past, right, for where you and me will spit in that first kind of 50k of our own money to kind of get the whole thing off the ground, then we might get a few kind of angels, get a few customers, might raise that 750k seed to kind of ramp it up, to kind of get to that, you know, 50K MRR and say, hey, this could be a real company.
And then we go out, raise a kind of a round or some version of that, right?
This was different.
Omer (16:53.890)
So if you didn't build an mvp, how did you spend that seed money to get to that thumbs up or thumbs down?
Guest (17:00.690)
So the very first piece of software that we wrote was not customer facing.
So if you think about what we're trying to solve here, which is some idea of us knowing what is being said in a conversation in regards to meeting scheduling, and that is not something where we're trying to solve language at large.
But we certainly need to know the intent.
As in, is then he's trying to set up a new meeting, cancel a meeting, reschedule a meeting, is he running late, is he making you optional me mandatory, extended the duration, and so on and so forth.
All of the things that can happen in my universe.
So the first thing was certainly one for we need to be able to define that universe in full so that we have an exhaustive list of intents and entities that are truly mutually exclusive.
Because if we can't even do that on the whiteboard, then I have no understanding of my universe.
Like, I can have no aspiration to go collect a data set that maps to that universe.
As in, I can't even go label anything because I'm not even sure what, what I'm supposed to label slash, annotate.
So that was the very first task, because if that turned into one of these kind of long tail type outcomes for where anything which you add to the list really requires another two things to be true, and those two things require another two things, then you end up with these kind of whole world problems.
And for us there was a little bit of a risk where we actually didn't know and we believed that we were able to, to come up with a reasonable finite list or some sort of drooping tail perhaps.
And on that list we built that initial kind of annotation guideline on if we can collect some data, how will we label that against these intents and entities?
And then we built an annotation console where we could hire some workers on that incoming stream of emails to then label them.
And that was an internal piece of software and that was what we ran on for the first period of time, even when we had a kind of human schedule just to kind of see can we even define the universe and then label it and then see whether those labels are an accurate representation of what is actually going on.
So that was that first almost year, and the outcome of that year was then at the end of it, we now know everything which is going on.
We can now see that we can define and label it.
And it certainly looks like if we get this amount of data, we can make this type of predictions at this level of accuracy.
That was not certainly something we knew for sure, but we certainly felt confident in suggesting that that would be the case and have some idea of then what would we need now to go build an mvp, which was our A round?
Omer (19:52.060)
Got it.
Your footnotes, by the way, have been very insightful.
So I think in the show notes I might create a footnote section with your footnotes.
So keep them coming.
Okay, so the seed round was about $2 million, right?
Guest (20:07.899)
Correct.
Omer (20:08.340)
Now, I'm not sure if there are a lot of founders out there who could raise $2 million for a seed round and say, I'm not going to focus on any customers, I'm just going to get to a point where I get a thumbs up or a thumbs down.
And I'm wondering, I mean, was your track record with startups a important factor in helping you raise that money?
I mean, you've had three successful exits, one startup that went bust, but do you think it was important for you to have that credibility to be able to go in and say, yeah, I'm just going to get to a thumbs up or a thumbs down with this $2 million.
Guest (20:46.220)
I agree.
It would be naive for me to suggest that the past outcomes didn't have some effect on our ability to create this particular narrative.
And certainly any entrepreneur with any exit receives an entry card into some tiny club of people who created something from nothing and turned it into an amount of dollars on your other side, however dramatic or however tiny, probably doesn't even matter, just that you were able to take it from the beginning to the end.
So certainly agree on that.
I would like to believe though, that that's only one part of the equation.
I think the other part Is that too many of us, and I'm certainly guilty myself, will easily gravitate towards the non dramatic scenarios, the all or nothing, live or die, most likely die, and then pick a deck and presentation that is less of that.
But when you think about it, suddenly any traditional VC fund will probably want more people to come in with more dramatic stories.
But I think us entrepreneurs might just do ourselves a disservice by not presenting them.
Doesn't mean that we're all going to be successful in kind of delivering on that.
But I do think really leaning in is certainly more aligned to the very business model of the very people that we're pitching.
So I do think people should at least think about am I even making the ask big enough?
Omer (22:39.890)
Yeah, that's a good insight.
So I'm curious how you decided what your target market was going to be.
This is potentially a solution that could work for, you know, a consumer or you could go out and sell this to enterprises.
And in many ways I think you're doing both right now.
So like how did you kind of think about the market and segmenting that and where you were going to focus when you started?
Guest (23:09.710)
So we certainly believed that this would be a business productivity tool.
So when you suggest that we sell to the consumer, I think both you and me would then translate that word into a single individual who happens to have a job and is probably using this particular tool to schedule meetings for their job and not to kind of set up drinks, appointments or dates, correct?
Yeah.
So in that regard it's not a traditional consumer, it's a single business individual who might have gone rogue.
Just like you might have been one of the early adopters of Dropbox and you signed up for it because you need to share a big ass file and your company didn't allow you any way to do that, so you just did it yourself.
So you were in that sense certainly a business customer, but a single individual business customer.
So we knew that we would be in that particular segment.
We also knew that if we looked across that segment, each and every single knowledge worker would conceptually be a target of ours.
That doesn't mean that you should focus on all of them to begin with.
And we could certainly see that people who did one more external meetings for you have very little visibility into the availability of the other participants is a good segment.
Two people who do many meetings, by that very count, just in more pain, say recruiters, salespeople and so on and so forth.
So we certainly started out by leaning into those segments, but we can see that we kind of look like a Trello or a Dropbox for where people very quickly kind of self select, as in I do meetings, I don't like it, I'll go sign up.
What I also think we have here is this kind of opportunity to go to market in this new, newish bottom up way for where you can attack the enterprise not necessarily through the cio, but through the individual in pain.
And as a good example, even my own little startup here, Slack, never called me.
So how did I end up paying Slack $15,000 a year if they didn't call me?
Because Chris, one of our backend engineers was in a group for where they needed to kind of chat.
So he just signed up for the free edition for Slack, started using it, hit that ceiling of I think 10k messages, said hey Dennis, we're kind of using this.
Can't you just get access for all of us?
Sure.
And then before I know it, you got 80 people on that kind of contract and Slack had Chris be my kind of sales contact and then all of a sudden that contract was in place.
That is interesting.
I think we have some of the very same characteristics where we see people say from Microsoft or Uber or the New York Times or VMware Pepsi just sign up.
I don't have contracts with those companies.
I just have individuals in pain and then those individuals turn into little mini pockets where oh then Kevin's team started using it.
Oh then another team started using it.
And that should probably become the catalyst for me to kind of set up the more traditional enterprise sales where one day I can call on them and say, hey, you know what, you are already paying me.
How about we make this relationship a little bit more formal and get an SLA in place and do all of the things you should do if you want to serve kind of a big enterprise.
Omer (26:40.280)
And the product is really has some great sort of inherent virality in it as well that you know, if you're an employee at Uber and you sign up and you're scheduling meetings with your colleagues and you're copying your assistant, Amy or Andrew, people are going to be asking you questions in terms of what you're doing or how you got that assistant.
So then before you know it, everyone's using it.
Guest (27:04.920)
Yep, you're absolutely right.
And more than half of all the signups that we get have touched Amy or Andrew in some way, shape or form.
Meaning that by far the biggest acquisition channel that we have is the product itself.
Omer (27:20.200)
That's awesome.
Now the pricing model right now.
So if you're an individual, you can start using x AI $8 a month and then you have team and enterprise plans which kind of the enterprise version is like $24 per seat per month.
You used to have a freemium model at one point, but you told me when we were talking earlier that that didn't work.
So can you tell us about like what process you went through, what you learned from that experience?
Guest (27:51.720)
So if what I said before is true, then I should be super hungry to just see as many people as possible sign up because the more who sign up, the more exposure we will see, the more likely it is that we'll sign up the next guy.
So we certainly leaned into the idea of Freeman early on but I think the mistake we made or where we were perhaps a little bit naive to be honest was on the fact that we actually introducing a slightly new type of product for anybody and their mom will have some good idea of how to use the next iPhone app because they now been kind of educated in what to expect if I install an iPhone app or if I go to a web service in my browser.
Not everybody certainly at that particular point when we ran these initial tests and we brought it to market and everything.
And the premium model at that point was just one for where the first five meetings is on me.
If you want more, we got kind of a paid edition.
But it was one for where if people can't self onboard and self educate without any hand holding, your product is probably not ready for a freemium model.
Because in a non premium world I can dedicate people and resources to make sure that the people that we sell are well educated in how to use this.
So we just I think did it a little bit too early.
Of course today we're probably closer for where if I take a wild guess, you might even have some Alexa or Google home available to you.
If not that, you certainly tried say Siri on your phone over the last month.
So it's become a little bit more common to speak to your computer or this idea of interacting with an intelligent agent.
And I think we're coming closer to that point for where we can reintroduce the idea of freemium.
But we didn't see that back then and that was just a little naive.
So we simply couldn't make that work well and there was too much support coming along with those freemium users and we didn't get that loop going.
But I do think we're getting close to being able to do that again.
Omer (30:07.200)
So that's something you might test again.
Guest (30:09.760)
Yes, very much so.
Omer (30:10.960)
Now, earlier you said, we talked about kind of, hey, is that all it does in terms of having this assistant that you can kind of copy and ask them to do things?
Hey, is Dropbox just this?
You know, it just lets you save stuff in the cloud.
But there's a lot more complexity behind solving this problem space.
And that's, I think it's reflected in the fact that you've raised so far $44 million, right?
Guest (30:41.230)
Correct.
Omer (30:41.870)
And how many engineering people do you have working on solving this, improving this, like, tell us a little bit about so people can get a sense of the complexity and the challenges behind this.
Guest (30:54.510)
So just to give you an idea, if you take Jan 1, which would then be five years, about five years into the venture, you would have us run with about 70 people in Manila doing nothing but just label data and 50 people here in the office primarily doing engineering and data science.
We are now reaching the inflection point for where the labeling we do.
And you'll see many kind of products that are kind of built on the idea of machine learning reach some inflection point for where the sheer use of the product become the labeling.
I'm putting this in air quotes.
So right now let's take Waymo as an example for that self driving car, or really any other kind of machine learning exercise.
They are having a host of cars driving around with a set of safety drivers collecting data on those predictions and doing all sorts of kind of simulations on that, on kind of close kind of settings which they have, which is fine.
But at some point, once this is being deployed fully in society, the sheer use of those cars and things happening become the data input on how it's going to improve for tomorrow.
But you need to kind of figure out when do you believe you're ready for that?
We are a about to kind of hit that inflection point, actually, which I haven't talked about in public, but I think, you know, this podcast is as good as any.
So on June 1st, all of my kind of prior data labeling operation will cease to exist because I we've kind of reached that inflection point.
So we're super, kind of both proud and excited about and slightly nervous and all the things that kind of come attached to that because we spent so much, so much time in that basement labeling, what are we at now, 32,33 million elements to kind of get to this point where we now know that is so robust that any new labeling needed will come from its sheer kind of product use.
That's A major milestone where we just happened to talk today on it.
Omer (33:15.000)
So for people who aren't familiar with this, can you just describe what, what the labeling is?
So this is like when somebody sends an email.
Guest (33:22.380)
So the system receives an email or a message.
Really the message could be a slack message, an email, a kind of UI action.
So message is being received.
We need to understand what that message says.
And we'll obviously do all sorts of predictions on it.
But what we will also do is that we'll immediately disassemble it into small components and label those components.
So I need to kind of label the intent.
Say you're trying to set up a new meeting.
How did you say that?
Where did you say that?
And why am I labeling it as a new meeting?
Intent?
You might, in that new meeting or that same message suggests it's not just a new meeting, but you want to have it first week of June, in the afternoon for half an hour.
I need some labeling guidelines where I can both label the duration, I can label the temporal expressions, and we will have a massive labeling effort so that we can take those labels and you think of it.
That's obviously not how it works, but conceptually, as a piece of paper, you get five highlighters and a pen, you highlight all the elements.
Then with the pen, you kind of connect them on how they relate.
Because as you do that over and over again, that becomes the understanding we have so that tomorrow some point we can make very high accuracy predictions on exactly what it is.
You're saying that's kind of what people call supervised learning.
Omer (34:53.710)
Yeah, I mean, it's a fascinating space and a problem area that you're solving here.
I think we need to wrap up.
I know you need to run as well, so I'm going to kind of move on to the lightning round.
I'm going to ask you seven quick fire questions.
Just try to answer them as quickly as you can.
Guest (35:08.750)
Sure.
Omer (35:09.110)
Ready?
Okay.
What's the best piece of business advice that you've ever received?
Guest (35:14.290)
Don't quit too soon.
Omer (35:16.050)
What book would you recommend to our audience and why?
Guest (35:18.850)
Certainly for the entrepreneurs listening.
Go read the Narrow Road by Felix Dennis.
It's going to come off super rude and not very kosher, but the idea of ownership is strong in this book.
Omer (35:36.200)
What's one attribute or characteristic in your mind of a successful entrepreneur?
Guest (35:40.600)
Optimism.
Omer (35:41.960)
What's your favorite personal productivity tool or habit?
Guest (35:46.040)
Email.
Omer (35:47.400)
What new or crazy business idea would you love to pursue if you had the extra time?
Guest (35:52.119)
Here's the funny thing.
I would probably run some 40 room hotel in the Caribbean where I could walk around in my three piece suit seven days a week.
But I'm not so sure it's a business idea.
It's probably just more of a fantasy of what I want to do at some point in life and I'm not sure I'm ever going to reach it.
Omer (36:12.430)
I love it.
What's an interesting or fun fact about you that most people don't know?
Guest (36:17.230)
I do these very long walks, so on Sundays I would typically I live on Wall street, which is the southern end of Manhattan.
So I'll take the subway up to the other end of Manhattan.
What is that, 20, 25k away.
Then I'll walk all the way back.
No iPhone, no podcast, no nothing.
Just me and my mind.
And I get to kind of put everything in the right boxes after three hours when I arrive home.
And then it's a good little exercise, you know.
Wow.
Omer (36:48.530)
Yeah, I've heard a lot of people talk about the benefits of walking.
And finally, what's one of your most important passions outside of your work?
Guest (36:55.810)
I did mull that over.
I like the fact that I've been able to turn exercise into a habit for where you'll see me in the gym five, six times a week and if you do it often enough, it becomes that habit for where it's not a chore, it's just a way of life.
Omer (37:17.710)
Awesome.
Great.
Well, Dennis, thank you for joining me.
I kind of feel like we need to find a way to continue this conversation at some point.
There's still so much to talk about, but I appreciate you kind of joining me and kind of helping share your insights and kind of the story so far.
Now if people want to find out more about X AI or give it a try, they can go to X AI and if people want to get in touch with you, what's the best way for them to do that?
Guest (37:47.540)
I am everywhere on the Internet under Dennis Mortensen, so go wild.
And if you know those two things, you can also email me.
So I am not shy to kind of participate in discussion or kind of cold emails.
Omer (38:05.450)
Awesome.
Thanks again and I wish you all the best.
Guest (38:08.890)
This has been fun.
We'll do a take two in eight months.
Omer (38:12.170)
I look forward to that.
Cheers.
Guest (38:14.410)
Cheers.
