Latest / Founder Talk with Alex Sheridan / This AI Agent Builder REVEALS the Secrets to Creating Business Efficiencies and Opportunities

Transcript

This AI Agent Builder REVEALS the Secrets to Creating Business Efficiencies and Opportunities · Founder Talk with Alex Sheridan

• 0:00
  Today I'm joined by Andy Sharma, founder of Flotagen, a company
• 0:03
  that builds and deploys AI agents for businesses.
• 0:07
  And here's why I think this conversation is so important
• 0:10
  right now. And I'm really excited to have
• 0:11
  you on is because every founder I talked to, including myself,
• 0:15
  we're trying to figure out where is the best place to use AI.
• 0:18
  Should we be doing AI automations, just automations?
• 0:21
  Do we need to fix our processes first?
• 0:24
  AI agents, should those be going in our business?
• 0:26
  Should we hold off on hiring to develop more AI in the business?
• 0:30
  And I think it's also one of those things where you can go
• 0:31
  down a rabbit hole if you're not familiar with this stuff or
• 0:34
  you're not an expert in this stuff.
• 0:35
  And you can honestly, you can build some cool stuff down that
• 0:37
  rabbit hole that's useful. But you can also waste a lot of
• 0:40
  time if you're not careful too. And so it's like, I think
• 0:43
  there's a little bit of anxiety right now going around just
• 0:45
  where people like it's moving so fast.
• 0:48
  We were talking about it right before that we we got on air
• 0:49
  here or live here is that it's moving at a such a rapid pace
• 0:53
  that you almost have a little anxiety of like trying to keep
• 0:56
  up. And tomorrow something new came
• 0:58
  out. And then it's like, you know, I
• 0:59
  got to adapt this. But do you though, you know, and
• 1:01
  so I'm excited to have this conversation.
• 1:03
  Andy, welcome to the founder talk podcast.
• 1:05
  How you feeling feeling? Great.
• 1:06
  Thanks. Man.
• 1:09
  Yeah, good to be back, man. Yeah, you've been here before.
• 1:11
  So. So some things we're going to
• 1:13
  get into today, just for the audience to set some context
• 1:17
  here. A lot of things, but these are
• 1:19
  some of many. The hidden use cases around AI
• 1:22
  and AI agents that founders don't yet know about but
• 1:25
  desperately need to. Who should not be messing with
• 1:28
  AI agents yet? Who and who absolutely should
• 1:31
  be? What AI means for hiring
• 1:33
  headcount, how you grow a business going forward.
• 1:36
  We both have kids. We were talking about, man, what
• 1:38
  does this world look like in terms of their future?
• 1:40
  Because it's and even ours, but but definitely theirs are going
• 1:43
  through college or they're getting ready to go to college
• 1:44
  or maybe they're not going to college or whatever they're
• 1:46
  doing. And then how founders who get
• 1:49
  this right are building businesses that are starting to
• 1:51
  kind of run without them having to operate every single part of
• 1:55
  the day-to-day. And that's ultimately part of
• 1:56
  the mission that we're trying to accomplish here.
• 1:57
  So I want to say this too, and this is this has never been done
• 2:00
  before on the show. AI prepped this entire podcast
• 2:04
  episode. So I had an automation and AI
• 2:06
  automation set up. And this is going to be seem
• 2:08
  very basic to you that you do much more complex and cool stuff
• 2:11
  than this. But someone books to be on the
• 2:14
  show, they automatically get sent a video e-mail.
• 2:17
  There's some automations there. And then to create that guest
• 2:20
  experience, you know, and then what happens is then they get
• 2:22
  put in our CRM. They're LinkedIn URL, company
• 2:26
  URL filled out from the form that you filled out.
• 2:29
  And then AI goes through. Now it's clawed embeds in our
• 2:32
  system through Zapier integration and it goes through
• 2:35
  and preps. But here's the key.
• 2:36
  It's not just like, hey, prep for the podcast or I give it a
• 2:39
  few frameworks or a few templates.
• 2:41
  This is a long, exhaustive framework and I've I've given it
• 2:45
  transcriptions from my show, so it knows how I communicate back
• 2:49
  and forth with guest. It knows kind of how I take the
• 2:53
  flow in the conversation and knows what to avoid and what to
• 2:56
  stay on course with. It challenges assumptions.
• 3:00
  There's a lot of cool things built in there and it's still a
• 3:02
  work in progress. But let's see as we go through
• 3:04
  here, you guys let me know is this, did it do a good job with
• 3:07
  this prep? We'll see, we'll find out.
• 3:09
  So here was the first thing that it flagged first, first segment,
• 3:14
  kind of the foundation AI agents.
• 3:17
  What are they and what do founders need to know about them
• 3:20
  and how are they different from just regular AI?
• 3:21
  Cause a lot of people are using clawed chacha petite but not as
• 3:25
  many are using actual AI agents in their business.
• 3:27
  Sure, great question. And I'm, I'm impressed, Alex,
• 3:30
  that you did that. I did the same thing, by the
• 3:32
  way. So I, I, I put your, the pod,
• 3:34
  the podcast name, your name, and it spit out exactly what the
• 3:37
  podcast is about and what I should be prepared with.
• 3:39
  Wow, I got some of those things and it gave me good suggestions,
• 3:42
  but I, I sort of picked a few things from experience that,
• 3:45
  that I've seen recently. So we'll chat about that too.
• 3:48
  So what are AI agents so, and how do they differ from, you
• 3:53
  know, your, your chat bots that are out there?
• 3:56
  So it all began with in sort of November of 20/20/22 when
• 4:00
  ChatGPT, right? So it's a watershed moment.
• 4:03
  The whole industry changed because it has been around for
• 4:05
  many, many years with, with ChatGPT.
• 4:08
  What they did was they, they came up with something called a
• 4:12
  transformer architecture, which is so Google came up with this
• 4:15
  paper called in 2017, a paper, paper called attention is all
• 4:19
  you need. And it really dramatically
• 4:23
  changed how AI was going to be built.
• 4:26
  So attention is really about, so you've got these large language
• 4:29
  models that are they, they basically, I'm sorry, shouldn't
• 4:32
  have done that. So they mimic the human brain,
• 4:36
  right? So they got neurons that are
• 4:37
  artificial neurons, and billions and billions of them.
• 4:40
  Now, are those made by the NVIDIA chips?
• 4:42
  Do they have a big part of that? So they so are the, the, the
• 4:45
  neural network is actually software.
• 4:48
  It runs on the NVIDIA chips. So it's, it's trained that's the
• 4:51
  hard, that's the hardware and it's trained by.
• 4:53
  So training is really important. So how you train is attention.
• 4:57
  So it's basically it's like in your in your head, there's a
• 5:01
  neuron that says, every time I see Alex, I recognize him.
• 5:04
  And that's how, and how you know that is because a neuron in your
• 5:08
  brain fires. And so that that's what
• 5:11
  attention is really, if you think about it.
• 5:12
  So that's what attention is all you need.
• 5:14
  And that was a transformer architecture.
• 5:16
  So the large language models are all about ingesting pretty much
• 5:20
  everything that's out there. So they absorb the entire
• 5:25
  Internet, every book out there. And that's how they understand
• 5:29
  what, what humanity has come to understand his knowledge.
• 5:33
  So that's the knowledge base. But that's, so that's the
• 5:36
  generative AI part of it, which it basically what it does is so
• 5:39
  a large language model like chat GPD, like Claude is designed to
• 5:44
  really, if you break it down, it's designed to complete a
• 5:47
  sentence or complete like a music note.
• 5:50
  So it, it's, it's designed to predict the next word that comes
• 5:53
  in a sentence, right? That's the, that's the
• 5:55
  intelligence. And you do that through training
• 5:58
  through NVIDIA chips and it's a really expensive process.
• 6:01
  So that's that's sort of the foundation of everything.
• 6:03
  Can I ask you a quick question? Sure sure.
• 6:05
  Is Claude now better than Chachi BT?
• 6:08
  Because wild migration over and I've even switched.
• 6:11
  A lot of people I know have switched and it does seem to be
• 6:13
  better. I did too.
• 6:15
  So and I so this is a personal opinion and not, not, not
• 6:20
  political in any way, right? Right.
• 6:21
  It's not because of the the DoD thing that happened.
• 6:27
  Enthropic actually broke away from open AIA while ago.
• 6:30
  Which makes Claude. Which which makes Claude right.
• 6:32
  So open AI makes ChatGPT. They broke away from them and
• 6:36
  they basically they broke away because they wanted to build
• 6:40
  safe AI. So they they actually even have
• 6:43
  something they called it's called mechanistic
• 6:46
  interpretation, which is basically they look into the the
• 6:49
  large language model extremely expensive and difficult thing to
• 6:52
  do to actually predict exactly where the decisions are being
• 6:55
  made. So that when so that there's an
• 6:57
  example out there and they drop it did this to where in a large
• 7:02
  language model, while their models basically started
• 7:05
  blackmailing an employee because it was fearing that it was going
• 7:09
  to be shut down. I don't know if you heard that
• 7:10
  story. I did hear that story.
• 7:11
  That's a real. Thing yeah.
• 7:13
  And then they, what they do is so it's almost like an MRI on
• 7:16
  your brain. So that's what the this
• 7:18
  mechanistic interpretation does. It basically reads exactly what
• 7:23
  neurons are firing when the large language model is thinking
• 7:27
  about doing something like blackmailing.
• 7:29
  Now why would it? Why would AI want to blackmail
• 7:32
  somebody? Because I thought it doesn't
• 7:33
  have motivation. So it doesn't have its, it
• 7:35
  doesn't, you know, have intrinsic like things that it
• 7:39
  wants to accomplish by itself. It needs to be told to do
• 7:42
  things. So that was the old story.
• 7:44
  Now it's changed, right? It's getting scary.
• 7:46
  It is, It is getting scary, right.
• 7:48
  So, so the setup was as in about sort of vaguely remember the
• 7:53
  story. It's about a year old and a year
• 7:55
  is, is is is a generation century.
• 7:58
  In AI time. Yes, so the story was they set
• 8:01
  up a, a, a sort of a fake setup where it was a fake company and
• 8:06
  this guy who was running the app basically emailed somebody and
• 8:12
  and it was he gave the impression that he was having an
• 8:15
  affair with a Co worker and everything was make believe.
• 8:18
  And and then he sends another e-mail saying I think I'm going
• 8:21
  to shut down Claude or, or whatever the the model was.
• 8:25
  And the model panicked and it basically sent him an e-mail
• 8:27
  saying you cannot do that. If you do that, I'm going to
• 8:31
  send an e-mail to everybody in the company saying you're having
• 8:33
  an affair. But if you shut it down, how
• 8:35
  would I do it? Well, that that.
• 8:36
  So he was saying, don't shut me down.
• 8:38
  If you do, here's what I'm going to do to you.
• 8:41
  But if it's shut down, how would it do anything?
• 8:42
  It wouldn't do anything, right. So and that's that's what it was
• 8:45
  afraid of. But but the newest version
• 8:48
  that's come out, I don't know if you heard about the I think it's
• 8:51
  called and there's so many models that come out every day,
• 8:54
  right? It's, it's another entropic
• 8:56
  model and it's, it is super, super advanced to where it's
• 9:01
  doing this security vulnerability.
• 9:03
  So it's called 0 day vulnerabilities in software.
• 9:07
  And it's, it's really, really. So they're not releasing it to
• 9:09
  the public. I think it's called Mythos.
• 9:11
  That that's that's it just came out last week.
• 9:13
  Is that the AI that's so good that they're scared to release
• 9:16
  it to the public because of what it can do?
• 9:17
  That that's the one, that's the one, right?
• 9:20
  So in that one actually it it escaped the IT it actually got
• 9:24
  out and it got so, so to answer your question that what if you
• 9:28
  shut it down, it replicated itself somewhere else.
• 9:32
  So you could. So I mean this is like terminal.
• 9:34
  I think I've seen this movie before.
• 9:35
  Exactly. It's it's it's going to close
• 9:37
  with that, right. So anyway, going back to your
• 9:39
  question. So I know I digressed in in a
• 9:41
  lot of different ways there, but so the generative AI is the the
• 9:46
  is the transformer architecture of the large language models,
• 9:50
  which is just responding to you. So when you talk to Claude,
• 9:53
  it'll basically say, Hey, what do you want?
• 9:54
  Alex? OK, Andy's coming in.
• 9:57
  This is what he's about. What is do you want to want to?
• 9:59
  It's it's kind of a conversation, a very intelligent
• 10:01
  conversation, but it's only a conversation, right?
• 10:04
  Agentic AI is the next step. So think of the large language
• 10:08
  model as the brain. The agentic AI solution is the
• 10:11
  is the nervous system. So you can actually do things
• 10:14
  with it now. So as an example, you could, you
• 10:17
  could have an agent that actually calls me up and say,
• 10:20
  Hey, Andy, do you have any questions about the, the
• 10:23
  podcast? Do you, do you know where to
• 10:24
  park? Do you do?
• 10:26
  Would you like to know more about Alex and end this week
• 10:28
  conversational, right. And it would actually, that
• 10:31
  wouldn't even be an agent, but it would actually say, would you
• 10:35
  like to go for lunch with Alex and I?
• 10:36
  I'm going to make a reservation for you.
• 10:38
  What kind of food do you like and make a reservation seat by
• 10:41
  the window, right? What, what, what whatever,
• 10:43
  right. So you could do a lot of things.
• 10:44
  I'll give you some examples down the road.
• 10:46
  So long winded answer but I I wanted to make sure I I got to
• 10:50
  the next question which I forgot what it was now.
• 10:53
  I think that pretty much, and I was kind of like, you know, the
• 10:55
  automation versus agents versus AI and you know, 'cause I think
• 10:59
  that's we have kind of compartmentalize.
• 11:01
  It was like automations is helping you do if X happens,
• 11:05
  then Y happens, that's like a simple automation.
• 11:07
  There's multiple steps. And then AI came out, large
• 11:09
  language models which have been out.
• 11:11
  AI has been around for a while, but in terms of the public
• 11:14
  really using it and getting to play with it in this kind of way
• 11:16
  was new when Chachi PT came out and then now we're in this new
• 11:20
  wave. Everyone's kind of used to
• 11:21
  talking to AI now. And I think if you're not using
• 11:23
  it in your business as they're like creative sidekick or
• 11:26
  intellectual sidekick, you're just kind of crazy.
• 11:29
  Like you're just wasting time, you know what I mean?
• 11:30
  If you're not using it day-to-day to day-to-day.
• 11:33
  But now there's this whole AI agent thing.
• 11:35
  And so it's that I think is the piece that some people are
• 11:39
  playing around with it. I've talked to founders like,
• 11:40
  yeah, I built an AI agent and you know, I've played with it
• 11:43
  and there's companies that help other companies do it, I think
• 11:45
  at an elite level. But but this is the conversation
• 11:49
  that I'm excited about and I appreciate you going down these
• 11:50
  rabbit holes. I think people want to know the
• 11:52
  back stories and the history and, you know, kind of shape.
• 11:55
  And I think one thing is crazy does that Elon Musk was part of
• 11:58
  open AII. Don't think a lot of people even
• 11:59
  realize that in the beginning stages.
• 12:01
  And then he obviously left because they had disagreements.
• 12:03
  So seems like Sam Altman's kind of in some in some shit right
• 12:07
  now. He is.
• 12:08
  So there was a New York article that came out this last weekend,
• 12:11
  I think, and they basically talked about basically calling
• 12:17
  him a liar. And then the whole so there was
• 12:19
  a whole debacle that happened about a year ago where actually
• 12:23
  more than that it was probably 2023 for when he was removed
• 12:28
  from the company by the board. And then I.
• 12:30
  Remember that? And they brought him back.
• 12:31
  They brought him, so he he was brought in back, but so they
• 12:34
  fired the board instead, kept him And he he's a remarkable
• 12:37
  guy. I mean, don't don't get me
• 12:38
  wrong, he's super smart, but there's, they're questioning a
• 12:42
  few things about him right now. And I, I don't have an opinion
• 12:44
  on that subject, but that that's why he's personally being
• 12:47
  attacked. I think.
• 12:48
  Plus this whole whole thing we talked about AI is killing
• 12:51
  software as killing jobs. And I mean, the younger
• 12:55
  generation, you know, at some point we'll talk about that too,
• 12:58
  because that's. Yeah.
• 12:59
  Well, that's a good point. Like do you think AI is killing
• 13:02
  jobs? Do you think it's taking over
• 13:03
  jobs or because I think there's there's two aspects of this.
• 13:06
  One is like companies are just not hiring as much because of
• 13:09
  AI. And then two, they're straight
• 13:11
  up replacing human beings with AII think we're more in the
• 13:15
  people are pausing on hiring humans and figuring out can we
• 13:19
  use technology or AI to do more with less versus waves of people
• 13:24
  let being let go because of a is now taking their job.
• 13:26
  But I don't know you're the expert.
• 13:28
  What do you think? So, so 100% I, I agree.
• 13:30
  So I'm not not an expert on that subject on a verse of AI.
• 13:32
  I'm still still thinking about that, right?
• 13:34
  Because it's, it is, it is scary out there.
• 13:36
  You and I both have kids, you know what, what are their
• 13:39
  future's going to look like? The I, I, I, I completely agree
• 13:43
  with you. So I think it's companies have
• 13:45
  always done this right. They've always used an excuse to
• 13:48
  say, hey, we're going to just go ahead and lay off a lot of
• 13:50
  people because Wall Street rewards that.
• 13:53
  So I mean, really, you mean you probably look at right.
• 13:56
  And, and I mean, you see the the spikes in the stock prices the
• 13:58
  very next day. Yeah, right.
• 14:00
  So, so I mean, and unfortunately a lot of companies just care
• 14:03
  about that and. I think a lot of companies, to
• 14:05
  be honest, are just bloated too. And some of the, you know, not
• 14:07
  that everyone deserves layoff, of course, but I think there's
• 14:10
  probably a lot of roles where it's like, what are you really
• 14:11
  contributing to here in terms of value?
• 14:13
  Completely so after COVID, everybody over hired like
• 14:17
  everybody, right and now they're realizing that they have far too
• 14:20
  many people. So that that's sort of one side
• 14:21
  of the coin. The other side, I believe is so
• 14:25
  if you look at plot code as an exemption, you you use Claude,
• 14:27
  you're impressed with it. Plot code is a whole different
• 14:30
  level, right So there. So you can just basically go in
• 14:33
  and say, instead of writing, you know, 50,000 lines of code, say
• 14:38
  I want to build AI don't know a satellite tracking system.
• 14:44
  And this is these are the specs, these are the type of
• 14:46
  satellites. This is the ground control
• 14:48
  equipment. Go, go do this for me and it'll
• 14:51
  do that for you, right. So, so that I think in software,
• 14:54
  it's it's, it's impacting people already, it's impacting new
• 14:58
  college grads. So I think there's 40%
• 15:00
  unemployment in computer science, which was supposed to
• 15:03
  be the, the safest degree out there, right?
• 15:07
  And suddenly those, those folks are saying, well, what, what do
• 15:09
  I do now? Because, and I mean, I know kids
• 15:12
  from UFI in our neighborhood, which is an amazing and their
• 15:17
  computer science program is so highly rated, they don't have
• 15:21
  jobs. So you one year later, right?
• 15:24
  If you're a software company, yeah, and you sell sort of a
• 15:27
  basic software, I think you are very vulnerable because people
• 15:31
  can just kind of replicate like the the code that can create the
• 15:34
  apps now with AI so easily. And not to say that you can, I
• 15:38
  know there's people listening saying, OK, you keep it.
• 15:39
  If you just create an app and then you launch it like you got
• 15:41
  to sell it. You got to you got to have a
• 15:43
  real product that's valuable to marketplace.
• 15:45
  You got to have a team to distribute it to, you know,
• 15:47
  service the clients, but to a certain extent, the actual
• 15:51
  software of it, it's going to be easier and easier to create
• 15:54
  these softwares. And so it's like, it's something
• 15:55
  I'd be thinking about if I had a software as a service, I'd be
• 15:58
  thinking, could somebody just create this tomorrow and then
• 16:00
  we're kind of obsolete or they're going to undercut us or
• 16:03
  whatever it is. Is that a good line of thinking
• 16:05
  or you think that's kind of fun? It's, I think it is, it's, it's
• 16:09
  good to be thinking in terms of will somebody come replace me
• 16:12
  tomorrow? Right.
• 16:13
  You should. Everybody should be thinking.
• 16:15
  That you're kind of always thinking that in business I am.
• 16:17
  You guys probably are too, right?
• 16:18
  Listening. Like if you're a founder, you're
• 16:19
  always like, someone's coming for me.
• 16:20
  I got to. I got to up my game.
• 16:22
  I got to basically evolve. I got to service my customers in
• 16:25
  new ways, you know? Yeah.
• 16:27
  And this, I mean, this time around, it's a paradigm chef,
• 16:29
  right? It's completely in a lot of
• 16:31
  different ways, right? But it is, it is definitely a
• 16:34
  threat. I will say though, it's really
• 16:36
  easy to, to create a demo or, or, or like a little app that
• 16:40
  works really well locally or works well for you and your
• 16:44
  friends and family. But when you productionize it
• 16:46
  which which we give it to all your customers and you suddenly
• 16:49
  have a million people hit it versus 5 people, big, big
• 16:53
  difference. So that's it.
• 16:54
  That's exactly. It right, right.
• 16:56
  Because I think AI is almost creating sometimes an
• 16:58
  overreaction. Do you think that's the case
• 17:00
  where I think businesses are sometimes overreacting to, Oh my
• 17:03
  God, we got to change this, This isn't this.
• 17:05
  And and it's like, just hold on, dude.
• 17:07
  It's not going to change like tomorrow.
• 17:09
  And you're still going to need some of the basic human type
• 17:12
  stuff that you have to deliver on, right?
• 17:13
  But it is important to evolve with the times.
• 17:15
  That's why I think there's a lot of anxiety right now because
• 17:17
  it's just like, what do we do? You know, what we did yesterday
• 17:21
  is outdated already. It's like it's just, but then
• 17:22
  it's like, but we don't want to go too much in AI and lose who
• 17:25
  we are. Like lose the best part of us.
• 17:27
  You know, it's like, I'm doing this in person podcast because I
• 17:29
  believe that human connection is going to be more important than
• 17:32
  ever before. And I think it's one of the
• 17:34
  smartest marketing plays that you can do because this is the
• 17:37
  stuff that's going to be left over.
• 17:38
  Everyone's going to build the right post on LinkedIn.
• 17:41
  Everyone has their doing their meetings, the notes transcribes
• 17:43
  it, it filters it in their AI post and then it's out there or
• 17:47
  they got AI video avatars. These are the conversations
• 17:50
  though, I think people are going to always want to listen to and
• 17:52
  if they don't, then we're all screwed anyways.
• 17:54
  So I think it's but, but again, even in our business, as much
• 17:57
  human connection and relationship building is that
• 17:59
  there is and realness about it. You better believe we're
• 18:03
  thinking about AI every single day and we're thinking, you
• 18:05
  know, AI takes these three cameras and it when I'm talking,
• 18:09
  it cuts to me. When you're talking, it cuts to
• 18:11
  me. When we're both kind of going
• 18:12
  back and forth, it cuts to both. AI does that in 57 seconds.
• 18:14
  They used to take someone an hour to go through.
• 18:16
  So it's like, those are the things that I think you look at
• 18:19
  the process and you say, what can we do to make this more
• 18:22
  efficient? And let, let's maybe dive into
• 18:25
  some of that. What, what do you think?
• 18:26
  Where can founders use AI agents?
• 18:29
  You mentioned a use case or two when you come into companies
• 18:31
  like what are some of the, let's look at more, maybe some of the
• 18:34
  simple use cases and then maybe some of the more advanced
• 18:37
  because if I'm a founder listening, I'm like, alright,
• 18:38
  where can this stuff really make an impact in my business?
• 18:41
  Right, So at the, at the highest level, I'd, I'd maybe put
• 18:46
  everything in two categories. One is 1 is efficiency AI, which
• 18:50
  is taking the mundane and boring and repetitive part of your work
• 18:55
  and, and doing that faster and better and, and cheaper with AI,
• 18:59
  right. So you just already talked about
• 19:01
  it, right? I'm coming to see you, you do
• 19:05
  the research, you do the, the outreach, all the stuff that
• 19:08
  you, I mean, this is what you want to do.
• 19:11
  So I'm, you know, I spend a lot of time in healthcare.
• 19:13
  This is operating at the top of your license.
• 19:15
  That's what they call it in healthcare, right?
• 19:17
  So this is, this is what you're brilliant at.
• 19:19
  This is what you should be doing.
• 19:21
  Only this right? Everything else should be done
• 19:23
  by. So that's your efficiency.
• 19:25
  AI. What I'm really excited about is
• 19:27
  opportunity AI, which is the next bucket, which is what can I
• 19:30
  do with AI to really magnify my business, my, my revenue of how
• 19:37
  I engage with my customers, right?
• 19:39
  That I think that is really where the, it gets really
• 19:42
  interesting because efficiency and, and I mean, even from a
• 19:46
  sort of from a human perspective, efficiency is about
• 19:50
  cutting jobs. Opportunity is about creating
• 19:53
  more wealth and opportunity for everybody, right?
• 19:56
  So if you now instead of, I mean, you already doing so well,
• 20:00
  you've got 100 podcasts that you've already done, right?
• 20:03
  What if you were operating at, at, you know, 1000 podcasts a
• 20:08
  month kind of thing because you, you really, and I'm, I'm just
• 20:11
  making these numbers up, right? That's really not possible.
• 20:15
  But if you were just doing only the things that you want to do,
• 20:18
  then you, you, you could be and then you'd be creating, you'd
• 20:21
  behave. You know, I've got this kid who
• 20:25
  will now work for me and then, you know, the whole team's going
• 20:28
  to work for you. So that, that creates more
• 20:30
  opportunity now. So I, I thought about, so I
• 20:32
  asked Claude before coming in there.
• 20:35
  So Alex does this podcast that you already know about.
• 20:38
  So it, it remembers, right? So it's got memory.
• 20:40
  And and he goes, yeah, that's tomorrow, right?
• 20:42
  I'm like, yeah, that's tomorrow. And I said, I'm going to go see
• 20:46
  him. What, what, what do you think?
• 20:48
  What, what, what kind of ideas? It laid out a lot of things in
• 20:53
  terms of LinkedIn research, you know, cutting up those videos,
• 20:56
  as you said, distribution to the right channels, this and that.
• 21:00
  And then he talked about the community.
• 21:02
  I'm calling him he Claude Claude dead right.
• 21:04
  And then I said I said what about opportunity AI and
• 21:08
  immediately understood what I meant there.
• 21:10
  And so, well, you know, you could maybe take advantage of
• 21:13
  the community, you know, basically do research.
• 21:18
  And then so it, it was an incomplete thought.
• 21:21
  And I actually had a client who did something like this a little
• 21:24
  while ago. And so I want to suggest this to
• 21:28
  you and see if this this made would make.
• 21:29
  Sense we love doing real workshop stuff on on the show.
• 21:31
  Absolutely. So this, this is some, this is a
• 21:33
  real life experience I've had with, you know, building this.
• 21:36
  So so it's, you know, you've got this community of founders,
• 21:40
  right? And the AI can go research them.
• 21:43
  So you already got connections and, and what they said on the
• 21:46
  podcast. So it's got a knowledge base
• 21:48
  already. Plus it's their, you know, their
• 21:50
  digital exhaust social media, you know what, what have you X
• 21:55
  and, and LinkedIn. So AI can figure out who they
• 21:59
  are as a, as a person, whether they like bourbon or wine, you
• 22:03
  know, you know those kind of things.
• 22:05
  And then you can say, Hey, you know what?
• 22:07
  I, I know this one guy who, when he came on the show, said, I
• 22:11
  want to start a new tech consulting company or a new tech
• 22:14
  product company. And I had another guest who's
• 22:17
  actually looking for APE firm, AVC firm was actually looking
• 22:21
  for that. How about if you, Alex, you say
• 22:25
  what, why don't I arrange a dinner for the three of us?
• 22:29
  I'll, I'll play the role of connector and I'm, I'm going to
• 22:31
  be your marketing arm. I'm, I'm going to be your
• 22:33
  virtual agency, right? And I want you guys to talk
• 22:36
  about what that business looks like.
• 22:38
  I'm going to help you on the business side of things.
• 22:40
  You guys talk about where the money is going to come from and
• 22:42
  what the product is. So, so, so you could create
• 22:45
  those and doesn't have to be just three people, right?
• 22:48
  It could be an event. You you say you guys are like
• 22:52
  minded based on my research. And this is a topic and, and
• 22:55
  again, it'll help you come up with the topic of interest,
• 22:57
  right? So, so things like that, that
• 23:00
  would be a whole new line of business you'd go after and you
• 23:04
  could have yeah. And, and, you know, give, give
• 23:07
  both of them sort of a private, sort of a closed loop, a private
• 23:12
  AI that they can talk to you about their desires.
• 23:15
  And, and, and then that gets documented and then that becomes
• 23:18
  a business plan, right? So they think things like that.
• 23:22
  You could, yeah. So you could, you could mean
• 23:24
  there's a lot, I mean, and I, I really want to do more
• 23:27
  opportunity AI. Unfortunately, most clients
• 23:29
  right now are all about efficiency.
• 23:31
  And we're doing that too. So we're doing that at scale
• 23:33
  that I I've never heard. Someone breakdown AI in that
• 23:35
  way, efficiency versus opportunity.
• 23:38
  I think that's such a good breakdown.
• 23:40
  So much valuable and what you said before, I hope you all were
• 23:42
  listening to this because what he said was so important.
• 23:45
  If you could extract this one thing from this episode, let it
• 23:47
  be that you said operating at the top of your license or what
• 23:51
  was the phrase that that is it. That is it meaning that if
• 23:54
  you're the doctor or the dent or whatever it is or therapist, you
• 23:58
  should be using the that to conversate and or have
• 24:01
  conversations with clients to make an impact with them.
• 24:04
  If you're a podcaster or you're a marketing agents or whatever
• 24:06
  it is should be working with clients or developing, you know,
• 24:09
  the things that you do hosting the podcast, whatever that is.
• 24:11
  Maybe it's being on stage for some people.
• 24:14
  And I think so many founders right now are in the weeds in
• 24:16
  the operational stuff. And it's like guys, y'all should
• 24:19
  be doing this stuff like getting your company out there.
• 24:22
  It needs to be exposed and you need to do podcasts, you need to
• 24:24
  put content out there. And so creating demand for the
• 24:27
  business as we say. Gosh.
• 24:30
  So here I had a question. So love that strategy that you
• 24:33
  just laid out because part of what we're trying to build here
• 24:35
  is not just the podcast, but the community.
• 24:37
  It's all local founders in Chicagoland.
• 24:39
  And I think there's we're putting on an event hopefully
• 24:42
  later this year, like a conference for local founders,
• 24:44
  entrepreneurs. And that's more to come on that
• 24:46
  guys. So that fits in line with what
• 24:49
  we're trying to do. So I appreciate you sharing
• 24:50
  that. My question is, how does AI know
• 24:54
  that that is a strategy? Like where is it getting that
• 24:57
  information from? Because people will tell you AI
• 24:58
  is not creative and it's like, Are you sure about that?
• 25:02
  Yeah. Because it seems like it's
• 25:03
  pretty. Damn creative to me, whether
• 25:05
  it's getting it, I mean opposites getting it from
• 25:07
  somewhere, but how did AI, where did it?
• 25:10
  Like, how did it know to put that together right?
• 25:13
  So I agree with you. Yeah, it's super creative.
• 25:17
  If if anything, it's, it's, it's probably the most creative thing
• 25:19
  anybody's ever created, right? The, and I, I come from
• 25:24
  consulting, spend 20 years with Accenture, right?
• 25:27
  And, and 10 years with other companies before that.
• 25:30
  The consulting world is in jeopardy right now that they're,
• 25:33
  I mean, every day you see some news, you know, so software and
• 25:37
  consulting are struggling. Why?
• 25:39
  Why is consulting? In trouble so.
• 25:42
  Let me know talk about my my motherhood, Mother Mother
• 25:46
  organization because yeah, just in general, I.
• 25:49
  Love them and it's too emotional.
• 25:50
  For me to you talking about that, right?
• 25:53
  Let's take Mackenzie. What, what do they do?
• 25:56
  They go in and they sell, tell clients, what are you looking to
• 26:00
  do? Here's we're going to in the
• 26:03
  next month, we're going to generate a strategy document for
• 26:07
  you. You can do the very same thing.
• 26:09
  Go to Claude, go to go to ChatGPT, go to Google Gemini.
• 26:16
  And I'm, I'm not, I'm not going to say much more about that, but
• 26:19
  you, you basically get in there and you type your prompts.
• 26:22
  You can even tell the AI to assume a certain personality.
• 26:27
  So you can so that there there is.
• 26:29
  So prompt engineering is pretty sophisticated now.
• 26:31
  So you can basically say you are a management consultant with 30
• 26:35
  years of experience. You are going to be, you are
• 26:38
  focused. You, you're really focused on
• 26:40
  XYZ industry. You're, you're sort of, you
• 26:46
  know, core competence is supply chain management.
• 26:51
  I'm, I'm just making all the stuff right.
• 26:53
  But you basically tell them exactly.
• 26:55
  So what you're doing when you you giving them a personality as
• 26:58
  you telling them where to go within their within their corpus
• 27:02
  of data. So how they're creative is
• 27:05
  they've ingested everything. So all these large language
• 27:07
  models have everything, be it that has ever written.
• 27:11
  So it's pulling from, not just the.
• 27:13
  Internet of Things, but it's pulling from conversations
• 27:16
  everything that it has with the people using the AI, which is I
• 27:19
  think it says it doesn't do that, doesn't it?
• 27:20
  So so it does do that. So it depends on how you
• 27:23
  construct it, right? So you can so they have
• 27:25
  enterprise license. So if you have Open AI Azure or
• 27:28
  enterprise license it, it prevents it from going back to
• 27:30
  right. There's a wall there.
• 27:31
  Yeah, right. But.
• 27:33
  If you just, if you and I are just using our 20 bucks a month
• 27:36
  subscription, it's going back. So it's essentially stealing.
• 27:39
  An idea for someone else? One 100%. 100% and then they
• 27:43
  what they did was they actually and I'm I'm not going to tell
• 27:45
  which company did this, but they bought every book out there and
• 27:49
  they that were not digital. So not all books were digital,
• 27:52
  right? And they basically took the
• 27:55
  pages out, scanned them, put another large language model
• 27:58
  that'd take a long time and there.
• 27:59
  Was a lawsuit about that. Right.
• 28:00
  And each, I think each author got like 30 bucks for each book.
• 28:05
  Wow. I mean, it was a class of action
• 28:06
  lawsuit. It was a, you know, a billion
• 28:08
  something dollars. But by the time it came down to
• 28:10
  the authors, it was nothing. Yeah.
• 28:13
  So, so they've got all these. So anything you can that's been
• 28:15
  ever written, that's and and I've got a funny anecdote for
• 28:20
  you, please in AI don't want to no, I'll come back to it.
• 28:24
  I want to I. Want to remind me to come back
• 28:25
  to the consulting strategy document that you're creating
• 28:29
  'cause I'm going to, There's a little bit of pushback I have on
• 28:30
  that. You may actually agree, but go
• 28:33
  ahead so. I'm not going to name the
• 28:35
  company, they they were a competitor, a big competitor of
• 28:38
  ours. And if you do the research
• 28:40
  you'll find him. So there was there was a
• 28:43
  consulting project with the government of Australia and the
• 28:47
  analyst on the team that's a junior most person in the
• 28:50
  consulting organization or consultant.
• 28:52
  One of the people that were I I guess that that was a story at
• 28:55
  least they used Chad GBT, right And Chad GBT hallucinated so
• 29:01
  and, and as it does. As they they.
• 29:04
  All do, right? And basically they presented a
• 29:07
  report and there's, there's actually a YouTube article on
• 29:10
  like a whole story on this. And they said, they cited
• 29:13
  sources in there saying in this lawsuit, this happened and
• 29:17
  that's why you should do this. Well, the lawsuit never existed.
• 29:21
  And they said there was a, a book written by this person and
• 29:24
  they entered and who claimed this is how you should do it.
• 29:28
  Well, the person who wrote the, who supposedly wrote the book
• 29:31
  was interviewing and she was, I didn't write the book.
• 29:33
  I wish I had, but I didn't. So they, they, they give their
• 29:38
  strategy document to the government of Australia.
• 29:41
  And then they basically when, when this was this all came out,
• 29:45
  they said, well, instead of said they building $400,000 back in
• 29:48
  the day, I'm sorry, and they said we'll give you $100,000
• 29:53
  credit or something. So it was it was just a really
• 29:55
  it was just a bad PR thing, which so, so, so they themselves
• 30:01
  are using large language models to come up with strategy
• 30:04
  documents. So, but, but I'll let you get
• 30:05
  get back to your question. Well, my thing about like, hey,
• 30:08
  because. I know there's there's going to
• 30:09
  be some consulting, there definitely is consulting firms
• 30:11
  and advisory firms, just you know, companies out there that
• 30:13
  do strategy and they coaching that kind of stuff, right, My
• 30:16
  push back. On.
• 30:17
  AI can now do it. Is that, you know, I was
• 30:19
  thinking, I was putting together a content strategy for a client
• 30:22
  recently and I'm, of course I'm using AI as I go through, it's
• 30:25
  like my intellectual creative partner, right?
• 30:27
  Helps me brainstorm some things, think through these things,
• 30:30
  process. It takes notes that I have in a
• 30:32
  meeting and distills it down to the important stuff.
• 30:35
  And I use it and I give it things.
• 30:36
  Let's organize this together. Let's do this together.
• 30:38
  But I thought about it like if I took myself out of the equation,
• 30:42
  yes, it could generate a content strategy, but it'd be like it
• 30:45
  could be way off from what this client actually needs.
• 30:49
  Now, you could argue if you gave it more inputs and you give it
• 30:51
  the right inputs, it could probably get there.
• 30:53
  But there's certain things that are like nuance that are from
• 30:55
  experience that, you know, I need to sit with this and work
• 30:59
  back and forth with the AI. It's not here's what I want,
• 31:02
  give it to me, boom, I've got it.
• 31:04
  And then there's the other piece, which is the execution
• 31:06
  of, it's like you have the greatest strategy document in
• 31:08
  the world. Some companies out there have
• 31:09
  the greatest strategy, but then it's like you need the coaching,
• 31:11
  the accountability, see the studio, you know, whatever you
• 31:14
  need to get your thing off the ground and executed at a high
• 31:17
  level. You know, as I sat in on a
• 31:19
  coaching session the other day yesterday with a client and they
• 31:21
  were recording videos and I was giving them real in time, real
• 31:25
  life feedback, You know, through what they were saying or how
• 31:27
  they were saying or where they were positioning themselves or
• 31:30
  how to make it more edutaining or whatever.
• 31:32
  It might be that I think this is the stuff that AI would not be
• 31:36
  able to do at this point or not as well as a human.
• 31:38
  So I to me, I think it's like, so I'm going to go back to you
• 31:41
  and get your thoughts on that. And then what does happen to
• 31:43
  companies like McKenzie or consulting firms?
• 31:45
  Do they just get smaller? Because now instead of needing
• 31:49
  1000 people, maybe you need 200 because you have the use of AI
• 31:52
  does help us do things a lot faster and more efficiently,
• 31:55
  Right? Right.
• 31:56
  So it's a great question and I, I agree with kind of your
• 32:00
  observation that it's not possible for you how to do that
• 32:03
  to the degree that you want it. The only difference I'll say is
• 32:07
  that's because you're using generic AI, right?
• 32:09
  So you're using something that's been trained on everything out
• 32:13
  there that that's, that's what hallucination comes from.
• 32:15
  Also, by the way, hallucination, believe it or not, is do you
• 32:18
  follow Geoffrey Hinton? He's the godfather of AI.
• 32:22
  So he's he, he started this whole artificial neural networks
• 32:26
  and he, he talks about the story where, and again, I'm
• 32:30
  digressing. No, please.
• 32:32
  The when when the Nixon trial was happening, there's there's a
• 32:39
  John Dean was his lawyer and he didn't realize that all the
• 32:45
  conversations were recorded. So when he was testifying
• 32:48
  against Nixon, he came up with lots of extra stuff that that
• 32:53
  didn't really happen because they're like, let's just play
• 32:56
  this. Yeah, exactly right.
• 32:58
  And he didn't, he didn't do it out of malice.
• 33:01
  He just didn't remember. And he just so it's a human
• 33:04
  thing too. I mean, if you, when you're, if
• 33:06
  it's, you know, a year later, you, you just make accept these
• 33:09
  things together. You, you.
• 33:10
  Yeah. So and that's, that's what
• 33:11
  hallucination is, if you really think about it, right.
• 33:14
  But it is possible. So if you were doing it at a
• 33:22
  well, I'll just call it production ready.
• 33:23
  So, so you, it's ready for business, you would build a
• 33:27
  different thing. So, and that's what we do,
• 33:28
  right? As a company, that's what we do.
• 33:29
  So we sit down with a client. I think what you're saying is
• 33:32
  extremely important. Very, very, very important too.
• 33:35
  And that's something I learned from Accenture.
• 33:36
  You got to understand what the client's trying to do.
• 33:38
  That's the number one thing, right?
• 33:39
  I mean, you, you can have a really fancy tool if it's not
• 33:42
  solving your problem, right? So really know how to execute
• 33:45
  it. Too, Not just what they want,
• 33:46
  but you got to know how to get there because it may give you
• 33:48
  shit that it's like this isn't going to work for a business.
• 33:50
  But even though it's the right answer, right, it's not right
• 33:52
  actually. So and that's.
• 33:54
  So what, what, so again, I'm, I'm going to quickly talk about
• 33:57
  what, what our model is because I think it's important and only
• 34:00
  because I want to, we, we were passionate about this, right?
• 34:03
  Because there's, you can do a lot of different things with,
• 34:06
  with, with tools. What we said, look what we want
• 34:09
  to do advisory research and engineering.
• 34:11
  That's what we want to do as a company advisory, because we
• 34:14
  want to understand what clients care about, right?
• 34:16
  There's a lot of noise in AI. Everybody's doing AI, right?
• 34:20
  And not everything can be solved with AI.
• 34:23
  We want to sit down with a client and, and we're very
• 34:24
  honest about that. If something is not an AI
• 34:27
  solvable problem, we'll tell them that we're not going to
• 34:29
  just make something up right. And as a part of that, we also,
• 34:33
  it's got to hit ROI. It's got to, it's got to hit
• 34:35
  your whatever metrics you have in mind.
• 34:37
  If it, if it's not hitting it, we'll just tell you and, and
• 34:40
  we'll, we'll gladly walk away because we don't want to, you
• 34:42
  know, build something that that doesn't work for you.
• 34:45
  So that's the advisory part of it.
• 34:47
  Research as we, we talked about, there's, there's been a Cambrian
• 34:50
  explosion of models, right? Every day a new model, new
• 34:53
  tools, open claw, you've got this, you know, new entropic
• 34:57
  model that's come out. So every day there's something
• 34:59
  new happening. So our, our, our team is PhD
• 35:02
  researchers. They actually spend, I mean,
• 35:05
  each one of them reads like a couple of papers a day.
• 35:07
  So really how do you build things?
• 35:09
  You know, so they're very good at that.
• 35:10
  And then engineering. So it's, it's the engineering
• 35:12
  again, very, very sharp. So we've got product people
• 35:15
  versus services people. And so they're they're
• 35:17
  accustomed to building products, but they're doing services work,
• 35:20
  which is a big challenge for us because they're not happy doing
• 35:22
  that kind of work, but they're really, really smart and they do
• 35:25
  it do it really quickly. So that that's kind of the
• 35:26
  model. So why I told you this whole
• 35:31
  thing is because once you understand what the client
• 35:34
  wants, it is possible to design a solution.
• 35:37
  So as an example, if if you were giving us clear directions that
• 35:41
  are maybe 17 pages long, that's how I say it's got to be, it's
• 35:45
  got to do this for me, It's got to do this for me.
• 35:47
  Then you figure out how to build that agentic AI solution.
• 35:50
  So you, you pick the right large language model or a combination
• 35:53
  of large. So a lot of our solutions will
• 35:54
  have multiple large language models.
• 35:56
  Use Llama parts for scanning, use Clod for more reasoning
• 36:01
  stuff, chat GPD for research. I mean, so you, you combine them
• 36:05
  all, right? And, and the agents basically.
• 36:08
  So there's, there's multiple agents also, and each of them
• 36:11
  knows exactly what their role is, right?
• 36:14
  And that agent will satisfy your different requirements, right?
• 36:17
  So each, each agent has a responsibility for.
• 36:20
  And then when you put it all together, it does become a
• 36:22
  custom solution that, so we used to say this at Accenture quite a
• 36:25
  lot. Your client and you as the as
• 36:29
  the consultant one the same thing.
• 36:31
  It's a solution that exactly meets the client's needs, right?
• 36:34
  It meets your exactly meets your needs.
• 36:37
  Otherwise it you know, what's the point, right?
• 36:39
  So yes, I mean, so with the 20 bucks a month, you'll get very
• 36:43
  far, but not far enough. With a custom solution you
• 36:45
  actually can accomplish what are you looking to do?
• 36:48
  So what's the big difference? I mean, are you guys just
• 36:50
  spending more time training? There's not like they're still
• 36:52
  the same LLMS there, right? Same brains, but you're just,
• 36:56
  are you spending more time on the training aspect on the front
• 36:59
  end to get to put more of the company in the AI or what's the
• 37:01
  difference between that and me going to Claude or ChatGPT,
• 37:05
  right? So training and this is probably
• 37:11
  again, a little bit more than I, I probably should be talking
• 37:13
  about this because it's it's too, too lengthy, but no one
• 37:17
  listen right now. Everyone tune this out.
• 37:18
  All right, so. Training is a really expensive
• 37:21
  thing. So training is what these
• 37:22
  companies are doing. That's why they're spending
• 37:24
  billions and trillions, right? Because that's what their role
• 37:27
  is, right? Our role is to take that that
• 37:30
  brain that always super smart and and make it focused in a
• 37:35
  certain direction. So as an example, so there's
• 37:38
  something called, there's a process called grounding, which
• 37:41
  is basically taking that brain and grounding it in a specific
• 37:44
  thing. So podcast prep, let's say.
• 37:48
  If you do that right. So what's the number one thing,
• 37:50
  your personality, right, your mission as as a podcast provider
• 37:55
  or I'm just making, I don't know what right.
• 37:58
  And your mission is to how do you grow a business when when
• 38:02
  you walk away or you know, you know, so that the the nuance of
• 38:08
  exactly how you build a podcast, both from a, you know, sort of a
• 38:13
  philosophical perspective and from a tactical perspective.
• 38:16
  All that goes into sort of the it's, I wouldn't call it
• 38:20
  training because training implies you're, you're
• 38:23
  retraining the large language model, which you could do, but
• 38:26
  it will cost you a billion dollars, right.
• 38:28
  And NVIDIA chips and don't think I have a billion dollars.
• 38:31
  Later on, now that I think about it, exactly right.
• 38:33
  What we could do is, is figure out a way to educate the, the
• 38:39
  app, which, which is a combination large language model
• 38:41
  plus your agents about your specific needs.
• 38:44
  So the way you do that is you ground it so that there's a
• 38:46
  technique called retrieval augmented generation, right?
• 38:49
  Rag, rag. What that does is it basically
• 38:52
  takes your, your documents, right, and creates A vector
• 38:59
  database. I mean, it's the way it does it,
• 39:01
  but what it does, it is basically understands
• 39:04
  semantically what you're trying to do, right?
• 39:07
  And every time that you try to do something with a system, it
• 39:10
  actually inputs that information into the decision making
• 39:14
  process, puts it right under the context window, right?
• 39:16
  So basically it makes it yours. And this is how companies are
• 39:20
  doing it too, right? So we think about all the
• 39:22
  companies out there that say insurance company, here are the
• 39:26
  policies. This is exactly how we do
• 39:28
  things. Don't hallucinate on that or
• 39:30
  healthcare. This is how we're going to
• 39:32
  provide care for our patient. Don't hallucinate on that.
• 39:35
  What you do is you take something like rag and you put
• 39:38
  it in in the. So what it does is it basically
• 39:41
  says when it comes to the subject, do not go to the
• 39:44
  Internet. In fact, I mean, there's there's
• 39:45
  a company I'm I'm I've done a lot of work with.
• 39:48
  It's called Ben AI. So BNAI stocks on Fry right now
• 39:53
  in case somebody's interested. Good company.
• 39:55
  Yeah. So they, they've created a model
• 39:58
  that's sort of a private large language model that doesn't
• 40:02
  connect with anything. And it only they build models
• 40:05
  for things that Only Connect to client systems, right.
• 40:08
  So so it is possible to do that, if that makes sense.
• 40:14
  So what? Happens to McKinsey in the
• 40:16
  consulting firms. What are you coming back to that
• 40:18
  you know, if they have because it seems like their nest now
• 40:21
  doing AI stuff, you know, which is obviously part of it.
• 40:23
  But like, what do you think happens to an advisory, a
• 40:25
  consulting firm, a staffing and recruiting firm?
• 40:29
  What happens to these companies over the next few years, right?
• 40:32
  So again, opinion and and and and uneducated opinion because I
• 40:36
  I not not a not know nobody knows what's going to happen
• 40:40
  honestly, right I think. Their prices may come.
• 40:45
  Down. So it's it's not the so they're
• 40:47
  extremely smart people, Mckenzie's and I mean BCG's,
• 40:50
  these guys are incredibly smart people, right?
• 40:53
  Know what they're doing and the experience.
• 40:56
  So, yeah, so the, the LLM's will come up with the answer.
• 41:01
  You need that extra nuance that says, hey, I've, I've done this
• 41:03
  at another client and they did exactly this that that ULM is
• 41:07
  recommending and they failed. And here's why they failed,
• 41:09
  Right. So, right.
• 41:11
  So you, you need, you do need that experience.
• 41:13
  But I, I think the days of walking in and say, you know,
• 41:17
  I'm going to, I'm going to charge you 4000 bucks an hour
• 41:21
  just making it up. But I'm not 100 grand a month
• 41:23
  for. A strategy document?
• 41:25
  No, it's. It's in the 1,000,000, right?
• 41:27
  For, for those kind of things, right, right.
• 41:29
  Unfortunately, that's crazy. I think those days are probably
• 41:32
  over and, and they're probably reinventing themselves right
• 41:35
  now. I, I know one of the companies
• 41:37
  that I follow equal is actually calling a reinvention.
• 41:40
  Actually, it's, it's out there. If you, if you look up that
• 41:43
  word, you know exactly who's using it.
• 41:44
  But but they definitely are because they won't be in
• 41:46
  business. For very much longer if they
• 41:47
  don't, yeah. Yeah, yeah.
• 41:50
  So you think price, do you? Think AI is just going to drive
• 41:52
  prices for almost everything down, especially when we have
• 41:54
  like, you know, in the future of humanoid robotics, which I know
• 41:57
  are on on the cusp of doing some things now, but it's very early.
• 42:00
  But you got humanoid robotics building things or building
• 42:03
  houses or on factories. You know, they're doing shit in
• 42:06
  your house and dishes. It almost seems like AI will
• 42:09
  create a better standard of living for everyone if it
• 42:12
  doesn't build us all, If it doesn't take over the entire
• 42:16
  world, you know? What are your thoughts on that
• 42:19
  macro picture? I, I I.
• 42:22
  Think things will get cheaper I I, I mean it's already happening
• 42:25
  right so even even the subscriptions and all are coming
• 42:27
  down the problem is a lot of money is being spent to build
• 42:31
  these things so they're saying our power prices are going to go
• 42:35
  up because all these data centers yeah it is very power
• 42:39
  hungry very, very, very power hungry.
• 42:40
  I think there's, I believe like a Reg, just a simple search on
• 42:46
  any of the large language models is like 1000 * a Google search.
• 42:50
  So in terms of price, So it's some, and I can't remember
• 42:54
  these, it's my daughter was telling me the other day.
• 42:56
  It's like, you know, it's, it's like every time we do a search,
• 42:59
  like a bucket of water that's being wasted and I don't want to
• 43:01
  do that anymore. So, so there's, there's that.
• 43:04
  So I don't think those costs are evident to the pricing models
• 43:08
  are completely disconnected from how much money is being spent
• 43:10
  because the money is being spent by the, the big, big investors
• 43:14
  and that that are thinking about the opportunity that lies ahead.
• 43:18
  So I think there's a little bit of disconnect there.
• 43:20
  I don't know how they're going to recoup that.
• 43:22
  Plus you what's the point of having a lot of cheap things
• 43:25
  when you don't have a job? So, right, So unless if there's
• 43:30
  something that happens on the Ubi, so your universal basic
• 43:33
  income type stuff that that they keep talking about.
• 43:36
  And I think part of it is so and again, opinion, right?
• 43:40
  I think the AI companies are not doing a good enough job of PR,
• 43:46
  right? So there's so much fear out
• 43:48
  there in the marketplace because.
• 43:51
  So if you think about, you know, these pharmaceutical
• 43:55
  commercials, right, that you see on TV, it's the first 3 minutes
• 43:59
  or whatever 2 minutes is about the benefits, right?
• 44:01
  People smiling there. I was depressed before.
• 44:03
  Now I've got my life is better. And the last five seconds or 10
• 44:07
  seconds, they'll say, well, it can cause, you know, this, this,
• 44:10
  this horrible things, horrible things, right?
• 44:12
  Your kidney failure and, and, and all that and I think the AI
• 44:16
  companies are doing the reverse. Unfortunately they're they're
• 44:19
  doing it's this is this is very scary and and you know what it
• 44:22
  is maybe there's something they know that we don't maybe it is
• 44:25
  going to be scary, right. So I mean everybody out there,
• 44:29
  all these guys keep talking about, you know, you don't know
• 44:33
  what's coming. Everything is going to shut down
• 44:35
  and people are not going to have jobs and you know, it can might
• 44:39
  get away and take over the world.
• 44:41
  And they don't do enough about the opportunity.
• 44:45
  They don't I don't believe. And that's why I think people
• 44:48
  are attacking, you know, people personally, you know, because I
• 44:53
  mean, and they have a responsibility.
• 44:54
  I think if they're changing the world, you should deliver on the
• 45:00
  promising aspects of the technology also, right?
• 45:03
  And if it that doesn't exist, then we're all in trouble,
• 45:05
  right? But if you say this is going to,
• 45:08
  you know, improve so many things, well, what's it going to
• 45:10
  do for the people when, when robots are going to be, you
• 45:13
  know, 250 a month or whatever, right, $2.50 a month?
• 45:18
  What who's going to benefit from that?
• 45:20
  How am I as as a as a regular person benefit from that?
• 45:25
  So they they should probably maybe spend a little more time
• 45:27
  on that. There'll be a.
• 45:29
  Repair a robot repair guy, it's like not really that'll just be
• 45:31
  another robot. Like so actually, you know, I
• 45:34
  think this whole thing of like it's going to create all these
• 45:36
  new jobs. I mean to a certain degree, yes,
• 45:39
  but I don't think it's going to it's not going to be a one to
• 45:42
  one ratio. If we just think about the
• 45:44
  reality of it, it's not going to for everyone Job we lose, it
• 45:47
  creates one job. And people compare it to like
• 45:50
  industrial revolution and all these different Internet.
• 45:52
  And it was just different the the because the scale and the
• 45:55
  speed of how this thing operates and and it's intellectual.
• 45:59
  It was like that was the one thing that humans had for so
• 46:02
  long. It's like, OK, yeah, you can do
• 46:04
  a tractor. Yeah, your machine's stronger
• 46:06
  than me, right. Or you got a factory with a
• 46:08
  conveyor belt and you figured some process out that's more
• 46:10
  efficient than one person doing it.
• 46:12
  But you still can't think like a human.
• 46:14
  You still can't, you know, use your brain to logic or use
• 46:16
  creativity. And now it's like it can and it
• 46:19
  actually is way better in a lot of cases, right?
• 46:22
  I I think there's no. Dimension so I I completely
• 46:24
  agree with you, Alex. There's probably another
• 46:26
  dimension too, right? So when, when I don't know,
• 46:29
  when, when you were a farmer using just, you know, just your,
• 46:36
  your hands and your, whatever tools you could and, and, and
• 46:38
  suddenly, you know, suddenly it became highly mechanized and you
• 46:44
  found yourself not having a job. You could go to school and learn
• 46:47
  something and become a computer scientist, right?
• 46:51
  Or the, the problem this time around is there is no next level
• 46:54
  up. So you, you could, I mean, you
• 46:57
  and I could decide, hey, well, let's all become ports because
• 47:00
  that pays a lot of money. It doesn't, unfortunately.
• 47:03
  And and no, no, no disrespect to anybody writing poetry, but
• 47:09
  right, I mean, what, what would you do?
• 47:11
  I'm gonna paint stuff you. Know.
• 47:13
  Yeah. And I mean, how many Picassos do
• 47:15
  you have, right? Exactly.
• 47:16
  So you so you can't really an AI and AI is already better.
• 47:20
  They're only they're only, you know.
• 47:21
  You feed it everything. Picassos that are painted, it's
• 47:23
  going to make a painting line. So what are the things you
• 47:25
  think? Are going to matter what jobs
• 47:26
  matter the most during this new wave of AI as it takes, you
• 47:30
  know, takes out a lot of the mundane a lot of the repetitive.
• 47:33
  I think if you're an analyst right county, you know, if
• 47:37
  you're you know, potentially some aspects of coding software
• 47:41
  customer service elements to A to a certain degree.
• 47:44
  What are the things that it's kind of AI proof in your mind?
• 47:48
  And obviously I think. It's it's the.
• 47:50
  It's the interaction, right? With human to human interaction,
• 47:53
  I think that is that's always going to be, there's always
• 47:55
  going to be a need for this because it is different.
• 47:58
  I've done podcasts that were online and it's not, it's not
• 48:02
  the same, not the same at all. You know it's that.
• 48:07
  There's going to be so I was reading something recently, it
• 48:11
  said it it still makes sense to do math.
• 48:14
  Maybe not computer science, math is good, but humanities is even
• 48:18
  more interesting because, you know, you learn about and
• 48:22
  communication and, and it's, it's the softer EQ versus IQ is,
• 48:26
  is what people are saying. But I, you know, who knows you,
• 48:29
  you don't know. And, and that's something.
• 48:31
  So I wanted to talk to you about that.
• 48:32
  So I, I, I, I, I feel passionately about guiding
• 48:37
  people. And then maybe there's a,
• 48:38
  doesn't have any in the local community do this because it is
• 48:41
  getting fairly scary out there and, and there isn't.
• 48:44
  So maybe there's a group of people that can create a team
• 48:47
  that says, Hey, you know, kids or, or people that are losing
• 48:50
  their jobs to in corporate America.
• 48:52
  We'll, we'll help you. We'll help you with if it's OPS
• 48:55
  killing or if it's, you know, thinking about things
• 48:57
  differently there. There's got to be a group out
• 48:59
  there that's doing it because the, I don't think anybody's
• 49:02
  doing that right now. I don't think so either.
• 49:05
  Yeah, there's not too many. I mean, the the tech companies
• 49:08
  are are building great products. They should be thinking about
• 49:11
  that. But nobody is right now, right?
• 49:14
  Yeah, I agree. I.
• 49:15
  Think to me, it's like either you're doing AI stuff right or
• 49:19
  you're doing something that has a deep human connection element
• 49:23
  that just physically can't be replaced by AI, right?
• 49:26
  And if it is, then we're all screwed anyways.
• 49:27
  Like if this doesn't matter anymore then then we're all
• 49:30
  done, we're finished, right? Right.
• 49:32
  Or you're. Maybe doing?
• 49:33
  Something with your hands that is harder to do, you know, like
• 49:36
  you're building a house, but you know, not in the not too distant
• 49:41
  future there will be a humanoid robotic that's built in the damn
• 49:43
  house. But I mean, I think that's
• 49:45
  farther away than analyzing a spreadsheet and replicating that
• 49:49
  back. That's pretty easy for AI to do
• 49:50
  or think through a problem, a critical problem, or do research
• 49:53
  or gather information and distill it down.
• 49:56
  So I think those are maybe the maybe the three buckets is
• 49:58
  either you're in AI like you guys are and you're kind of
• 50:01
  leading the charge because it's just happening.
• 50:02
  It's not like, oh, I don't want it to happen.
• 50:04
  I wish we could go back. It's it doesn't matter.
• 50:06
  It's it's, it's already here. It's not going away.
• 50:09
  Even if America shut everything down, China and other areas, all
• 50:13
  these other countries, they're going to be doing it too.
• 50:15
  So we might as well be first in line, right?
• 50:17
  You either got AI deeply human connection, like if you put on,
• 50:21
  you know, local podcast in person type stuff, events, if
• 50:24
  you do retreats with people that really, you know, are some type
• 50:27
  of spiritual, you know, emotional intelligence type
• 50:30
  thing, I think I think would matter most.
• 50:32
  And then something with your hands.
• 50:34
  Those are kind of the three categories that you should be, I
• 50:37
  think 100% thinking. About and I'm really glad you
• 50:41
  brought brought the third category up because when he has
• 50:44
  these experts, I said become a plumber everybody says that
• 50:48
  right So yeah I mean I don't believe we're going to have
• 50:51
  robotic plumbers anytime soon I don't think anytime like in the.
• 50:54
  Next 10 years, yeah, I don't believe it's.
• 50:56
  Happening, but I think 20 years. Maybe, yeah.
• 50:58
  You know, right now, that's, that's the advice the AI
• 51:01
  companies give everybody, not me.
• 51:02
  I mean think how fast. Things are going it's it's
• 51:04
  almost inconceivable to think about 20 years the way how fast
• 51:09
  things are going. You know, and once you have that
• 51:12
  first. Major breakthrough with the
• 51:13
  humanoid robotics and it's like, I mean, they're out there now,
• 51:17
  do your research. They're in factories, they're
• 51:19
  doing shit, they're doing real jobs and they're starting to,
• 51:22
  they're slowly will be coming into houses.
• 51:26
  Figure AI's doing it. Tesla's got the bots that
• 51:30
  they're doing it's coming. But I do think it's yeah, I
• 51:33
  think it's going to be a while before they come in.
• 51:36
  That's a pretty complex thing to ask someone to do, to come in
• 51:38
  someone's house and like diagnose and then physically do
• 51:40
  things and risk to that. And it just, it just seems seems
• 51:45
  a little ways away. I I I think the the.
• 51:46
  Cost of doing that be to to too much do so that that's probably
• 51:50
  why they wouldn't one go in that direction.
• 51:52
  That's a great point because it's.
• 51:53
  Got to be somewhere. It's like, hey, if it cost us,
• 51:54
  you know, a billion dollars to do this thing, this one could
• 51:56
  have done for 100 grand. It's like is, is the met does
• 51:59
  the math, even math, you know, like it doesn't really make
• 52:01
  sense. So what are some of the exciting
• 52:04
  things that you're doing with customers that are creating ROI?
• 52:07
  Like if you think about maybe just a couple of examples, or
• 52:09
  maybe it's, it's things you've done in the past or things
• 52:11
  you're going to do in the future.
• 52:12
  Because if I'm again from a founder listing, I'm like, what
• 52:14
  are some of these things I could do in my business to, you know,
• 52:17
  create opportunity, create efficiencies and whether it's
• 52:20
  saving money, generating more revenue, creating better
• 52:22
  customer experiences, right, right.
• 52:25
  What are some of the things? So let me start with the.
• 52:27
  Past so I lost voice since 2022 before Gen.
• 52:31
  AI and even after Gen. AI, I did mostly conversation
• 52:34
  AI, which was voice based. So it's sort of intelligent,
• 52:39
  empathetic, bi directional communication, right.
• 52:42
  So you whether you're talking to a patient about their chronic
• 52:45
  condition and really holding their hand through the process.
• 52:50
  So I, I have a you know, and again, I won't go into my
• 52:52
  personally, but I, yeah, I was really meaningful in my life and
• 52:56
  teaching me something that nobody had the time to teach me.
• 52:59
  So doctors are busy, nurses are busy.
• 53:01
  They couldn't answer my question.
• 53:02
  Yeah, I did. And time permitting, I'll give
• 53:05
  you the example, right. The so that that sort of evolved
• 53:10
  into so doing conversational AI, using conversational trees
• 53:16
  before and then it became Gen. AI, which you forgot the
• 53:20
  conversation trees. The the large language modeled
• 53:22
  at all the thinking for you. We figured out a way to, to
• 53:25
  constrain it. I mentioned Ben AI.
• 53:27
  Those guys did a really good job constraining how, how the
• 53:31
  conversations happened so that you're not giving patients
• 53:34
  inaccurate information because that, you know, you could, you
• 53:37
  could kill somebody if you if you didn't do that, right.
• 53:39
  So, so did that. And then lately, So what we're
• 53:42
  doing with Flora Gen. is really workflows.
• 53:46
  So it's so I'll give you an example of one of our clients
• 53:48
  is, is in the business of processing insurance claims and
• 53:53
  they, they get all these claims from the whole world over,
• 53:56
  right, the whole country over. And basically they are.
• 54:03
  So each of their clients will scan every possible claim.
• 54:07
  I think I mentioned this to you the other day, put on a big PDF
• 54:09
  that's 20,000 pages long. You know, how do you like?
• 54:13
  So there's a human that actually looked through them and say, OK,
• 54:15
  this one needs to go, this guy, this one needs to go to this
• 54:17
  group, that sort of a thing. So we created a system that
• 54:21
  actually recognizes even handwritten things on there
• 54:25
  exactly what kind of a claim it is.
• 54:27
  And that automatically routes it to the right group.
• 54:29
  Save them like hundreds of thousands of dollars.
• 54:33
  So, and it was a journey we went through because you got to do
• 54:36
  sort of the initial solution. Then if that doesn't work, you
• 54:39
  got to do human plus AI, then you got to do learning and then
• 54:42
  you got to put it back in the, the context window, as I
• 54:45
  mentioned. So there's a way to do that.
• 54:46
  So, but there's a lot, again, a lot of efficiency AI.
• 54:50
  We do have people that are thinking about opportunity AI
• 54:53
  and what, what I'm really passionate about is is really
• 54:55
  the mid market. So the small and, and founders,
• 54:59
  right, because they're underserved.
• 55:02
  So either, you know, Fortune 500 global 1000, they've got the
• 55:05
  accentures and Deloittes and everybody else in there, right,
• 55:08
  knocking on their door. The, the smaller companies have
• 55:11
  never gotten that kind of attention, particularly if
• 55:13
  they're, and if you think about founders who don't want to have
• 55:16
  an IT shop, right? You, you just, we just go both
• 55:20
  efficiency and an opportunity. You go and say, Hey, what, what
• 55:22
  are you trying to do? How about so you know, you have
• 55:25
  the podcast example build, build, build a bunch of tools or
• 55:28
  if you've got somebody that's in the.
• 55:31
  I don't know. Retail business, right, help
• 55:34
  them create new avenues for revenue that that's sort of a
• 55:39
  thing, right. So that that's that's really
• 55:41
  where the interesting things are starting to happen right now.
• 55:45
  I love the example you brought up of.
• 55:46
  Calling the guest on the show and just saying, hey, excited to
• 55:49
  have you on the show tomorrow. Yeah.
• 55:50
  Do you have any questions? You have any?
• 55:51
  And here's some value add and yeah, it could be your voice
• 55:53
  that. 'D be wild.
• 55:55
  It's, I mean, so it's. Really easy to do that now.
• 55:57
  Really. Yeah, it's really easy to do it
• 55:59
  and you could do it yourself. What tools you could use to
• 56:01
  connect that. So all you got to do is get get
• 56:04
  11 labs. It's a really, really nice
• 56:06
  company, right? 11 labs that you can clone your
• 56:08
  voice. We can do it in 30 seconds or
• 56:10
  maybe a minute, I don't know. And it'll be like Alex is
• 56:14
  calling them and having an intelligent conversation about
• 56:17
  what to expect. And I can use clawed.
• 56:19
  Code to what would I need to? I think 11 levels is all.
• 56:21
  You need really. Yeah.
• 56:23
  It's. Got to connect to numbers so I
• 56:24
  need. The active campaign I'm using my
• 56:27
  e-mail or SMS I believe. It's got all that.
• 56:31
  If it doesn't give me a call and I'll I'll point in the right
• 56:33
  direction. But yeah, yeah, it's got, I
• 56:36
  think it's yeah, but it called for my number.
• 56:40
  I'm sure it. Can't so I'm a little fuzzy on
• 56:43
  that. So I was recently asked by one
• 56:45
  of my Co founders like why don't you build an app Andy?
• 56:47
  So I I say, you know what, I'm going to do it.
• 56:49
  And I started doing it and I stopped and I said not operating
• 56:53
  at the top of my license. That's not what I do.
• 56:55
  There you go. And I said, no, I'm not.
• 56:56
  I'm not going. To do that because I, I could,
• 56:58
  but I, I don't want to because it'll take me 10 hours to do
• 57:03
  what somebody else can do in 15 minutes.
• 57:04
  That's it. What's the point?
• 57:06
  And I don't want to learn that skill right now.
• 57:08
  Now is that? Something you would do for
• 57:09
  companies What I. Do what we just talked about.
• 57:13
  I wanted to do, of course. Yeah, of course.
• 57:15
  OK. And how do you work I'm?
• 57:16
  Curious just for people listening too, because I agree,
• 57:18
  I think there's certain cases where it's like, yeah, it's good
• 57:20
  to know how to do certain stuff, right.
• 57:22
  There's other cases where it's like, is this what I should be
• 57:23
  spending my time on or should I hire a company to do this
• 57:25
  professionally? Right, Right.
• 57:27
  How do you guys typically engage with IS?
• 57:30
  It yeah. Is it like hey I?
• 57:31
  Need this one project you do or is it like, hey, you know, a
• 57:34
  monthly partnership? Is it, you know, the whole scope
• 57:37
  like I'm curious and I'm sure people are listening, are
• 57:38
  curious. Yeah, so.
• 57:40
  We, so me and my Co founders, we, we basically have been
• 57:45
  through situations where one product companies were not
• 57:51
  listening to clients a second, the PE company that was funding
• 57:56
  your previous company ran out of patients.
• 57:59
  So the company shut down the mall.
• 58:00
  So we said we're going to do a few things.
• 58:02
  We're going to self fund, right? And then we're going to give
• 58:06
  clients an easy way to work with us.
• 58:08
  The self funding takes care of how much we need to charge
• 58:11
  clients. We we go in and and talk to a
• 58:14
  client and do a discovery. Could be a phone call, it could
• 58:17
  be a 2 day session, right? That's free.
• 58:21
  But everybody does a pro bono, right?
• 58:22
  So not, not a big deal, right? Beyond that, instead of giving
• 58:26
  them a proposal, we'll, we'll create a product for them,
• 58:29
  working products. And in this case, it'll be
• 58:31
  Alex's voice doing everything. The only difference is going to
• 58:34
  be it's going to be hosted on our side of the fence.
• 58:37
  So it's hosted on our environment and we charge $5000
• 58:40
  for that. That's under our cost.
• 58:43
  And we do that because we want to make it easy for clients to
• 58:46
  work with us, right We. Also it's.
• 58:49
  It's a no risk type of thing for the client.
• 58:51
  So if for whatever reason they don't like it, it could be
• 58:54
  anything. Don't act like your hat, Andy.
• 58:56
  We walk away. Nope, Nope.
• 58:58
  And and we destroy. So the product we make is based
• 59:00
  on their protocols, their data. The product is immediately
• 59:03
  destroyed. Their data is returned to them.
• 59:06
  And, and I mean in concept, right, it's returned to them and
• 59:09
  it's it's never used again. And we part as friends.
• 59:13
  No, no, no worries at all. If they do like it, then we we
• 59:16
  charge the 5000 and then we figure out how to productionize
• 59:19
  it in their environment. So it, it costs between 20,000
• 59:22
  to 50,000 depending on the complexity to make it go live in
• 59:26
  their environment, meaning they own it forever.
• 59:29
  OK, Right. So it's, it's their system.
• 59:31
  If they want help in running it, we can do that for a very
• 59:33
  nominal fee. Yeah.
• 59:35
  And but it, it, it's basically something they own.
• 59:37
  And so that that's, that's differentiator.
• 59:39
  Yeah, yeah, I was just. Cuz I remember you, we talked
• 59:42
  about you, we talked about this. I think it's.
• 59:43
  A cool kind of. Model to bring up I think other
• 59:45
  founders are good ideas from it too as well.
• 59:46
  Yeah, yeah, so, so. So that, and then one of the
• 59:50
  things we do, so especially with founders, what we do is we'll
• 59:54
  invest in building the product and we'll do Rev share.
• 59:57
  So you know, whatever the so we'll, we'll figure out exactly
• 1:00:01
  what it cost to build it. And then we'll we'll come up
• 1:00:03
  with a model that says, hey, you can charge $0.25 a call for
• 1:00:09
  that, but give us $0.03. Right, something like that.
• 1:00:13
  So do ref share and and we recoup and those are the kind of
• 1:00:18
  things. So you, your your main focus is
• 1:00:20
  how do you build a business and walk away or we'll build your
• 1:00:23
  technology solution, not not you the the founders, right.
• 1:00:27
  And and you as well. Yeah, yeah, I got you.
• 1:00:30
  But basically that will enable them to AI will do the work for
• 1:00:35
  them. So it's kind of passive income
• 1:00:36
  that's coming in as as a revenue stream or a replacement of what
• 1:00:40
  they do today. Wow, yeah.
• 1:00:42
  What? Are you telling your kids?
• 1:00:43
  Right now you have a kid in college and then a kid that's
• 1:00:46
  working, doing consulting, right?
• 1:00:48
  What are you telling them in terms of what's happening and
• 1:00:50
  what they should put put in their time, effort and skills
• 1:00:52
  into so I? I tell them what I didn't do
• 1:00:55
  myself, which is learn. Learn how to do that.
• 1:00:57
  Get familiar with the tools. Pay attention to what's
• 1:01:00
  happening out there. You know, you know, what's how
• 1:01:04
  is work being done? Because that's the future,
• 1:01:07
  right? So if you're in consulting, if
• 1:01:08
  you're doing any, so the people. So there's, there is, there's a
• 1:01:12
  quote out there and there's the famous journal who said, and I
• 1:01:14
  think I put put on my LinkedIn too, that if you, if you don't
• 1:01:19
  like change, you're going to like irrelevance even less,
• 1:01:24
  right? So, so, yeah.
• 1:01:26
  So you can either say, Oh my God, everything just changed on
• 1:01:28
  me or Hey, let, let me be relevant and let, let me get
• 1:01:32
  into this, really learn how to do it.
• 1:01:34
  Now whether they listen to me, I'm, I'm sure my kids are going
• 1:01:36
  to listen to this at some point. I don't know if they're going to
• 1:01:38
  listen to me or not, but that's what they should be doing.
• 1:01:40
  And. When you guys are designing your
• 1:01:43
  business, are you, are you mapping out kind of an org chart
• 1:01:46
  and saying human, human, Yeah, agents, AI agents, AI agents?
• 1:01:50
  You think that's the kind of the future of business?
• 1:01:51
  Is that what you're trying to do is eventually, you know, will AI
• 1:01:54
  be able to run 80% of the business you think?
• 1:01:58
  So we we are. So we're a young company, we've
• 1:02:00
  been around for 8-8 months now. So we, we're very tactical.
• 1:02:04
  So we'll say, OK, this is what needs to be done.
• 1:02:09
  This is what is human top of the license kind of work.
• 1:02:12
  Everything else should be AI. And we, we sit down and, and
• 1:02:15
  figure out how to even internal operations that that's what we
• 1:02:18
  do, right? And we're fairly, fairly
• 1:02:20
  successful at that. So I'll give you an example.
• 1:02:22
  Even in our, as we build systems for clients, we take pride in
• 1:02:26
  the fact that we we deliver AT50X velocity of anybody else
• 1:02:30
  out there. Reason we're able to do that is
• 1:02:32
  because we've got 30 engineering agents that are home grown.
• 1:02:37
  So that we we have agents working for us that do a
• 1:02:40
  majority of the work. Our engineers just go in and,
• 1:02:43
  and basically do the, the fine tuning, the top of the license
• 1:02:47
  kind of work that, that makes it a real, you know, good product.
• 1:02:51
  But our, our agents do most of the coding and it's, it's, it's
• 1:02:55
  different, like it's different from cloud code.
• 1:02:57
  These are custom agents that do that.
• 1:02:59
  You know, you take a phone call and, and you, you create a
• 1:03:02
  requirements document from that, you create an architectural
• 1:03:04
  blueprint from that and you create code from that, just from
• 1:03:07
  the phone call itself. So it's, it's those kind of
• 1:03:10
  things that, that we've already automated a lot of that stuff we
• 1:03:15
  do. So some clients will insist on a
• 1:03:17
  proposal. So even though we say we'll just
• 1:03:19
  give you a product, they'll say no, we'll give you a proposal.
• 1:03:22
  In the beginning we said no, we don't do that.
• 1:03:24
  And they said no, you have to do that.
• 1:03:26
  So we'll use cloud for a proposal.
• 1:03:28
  We'll just a statement of work. We'll just put it in there and
• 1:03:30
  say, hey, generate a statement of work comes out looking very
• 1:03:33
  nice and you just give that to the client.
• 1:03:35
  So that saves hours and hours. Well, Andy, this has.
• 1:03:40
  Been an incredible conversation, probably one of the best
• 1:03:42
  conversations I've had in the last six months.
• 1:03:44
  That's great. Honestly, I really enjoy, yeah,
• 1:03:46
  enjoy your company. And I just, I think we're
• 1:03:48
  talking about things that hopefully you guys listening,
• 1:03:51
  you feel a little bit maybe at ease or maybe excited to go and
• 1:03:53
  do the next thing or maybe to the opportunity versus
• 1:03:56
  efficiency, the three different buckets.
• 1:03:58
  We talked about the real life use cases.
• 1:04:00
  Hopefully it was extremely, extremely valuable for you guys.
• 1:04:03
  Where can people find you, connect with you, learn more
• 1:04:06
  about what you guys do so LinkedIn and.
• 1:04:09
  Then the name of the company is Flora Jan.
• 1:04:11
  So those two avenues would be and I'm I'm in Chicago land, so
• 1:04:17
  happy to meet with anybody and we'll be there at your event
• 1:04:22
  when you do this. I would even love to.
• 1:04:23
  Have you speak, if you're open to it at the event, we can talk
• 1:04:26
  more details. We're kind of in the planning
• 1:04:27
  stages right now, but I think it'd be we don't have anyone
• 1:04:30
  that is a specialist in what you do in AI and bring the whole
• 1:04:33
  kind of innovation technology side of things.
• 1:04:36
  I think it might be a full opportunity.
• 1:04:37
  Yeah, that'd be great. The last thing I want.
• 1:04:39
  To end on is I did prep with AI. If you guys are curious about
• 1:04:42
  how did it go and what do you think?
• 1:04:44
  I don't know. I think it did pretty good.
• 1:04:45
  I also will say this though, is I really, it kind of gave me the
• 1:04:49
  start of the conversation After that, it was pretty
• 1:04:52
  conversational. It was just very back and forth
• 1:04:54
  and so I really didn't rely on my notes a lot.
• 1:04:56
  Now this is after doing almost 100, actually probably over 100
• 1:04:59
  episodes now. So I think you get a feel for
• 1:05:02
  how to run the conversation a little bit better or your first
• 1:05:04
  20305075, you're very much relying on your notes.
• 1:05:09
  And I also maybe just really interested in the conversation
• 1:05:11
  too. And it helps when you have
• 1:05:12
  someone sitting across me that you're genuinely interested in
• 1:05:14
  too. But AII think if you were just
• 1:05:17
  starting off or hadn't done a ton, really helpful in terms of
• 1:05:20
  the prep and gave me some things to think about.
• 1:05:23
  So you guys let us know. Good to hear, what do you think?
• 1:05:26
  All right, thanks for tuning in guys.
• 1:05:27
  Andy, thanks again for being here.
• 1:05:28
  Appreciate y'all watching the show and we'll see you on the
• 1:05:31
  next one. Great.
• 1:05:32
  Thank you, Alex.