Latest / AI for Founders with Ryan Estes / The uncanny valley is real

Transcript

The uncanny valley is real · AI for Founders with Ryan Estes

• 0:00
  Most founders say they want scale.
• 0:01
  What they really want is control.
• 0:04
  Perfect funnels, perfect hires, perfect customers, perfect
• 0:07
  market timing. If your business only works in
• 0:09
  perfect conditions, it's not intelligent.
• 0:12
  It's fragile. The moment the market shifts, an
• 0:14
  AD account gets banned, competitor undercuts you, a key
• 0:18
  employee leaves, a new AI model drops, everything shakes.
• 0:21
  That's not scale, that's staged stability.
• 0:24
  Real intelligence adapts. So are you building something
• 0:27
  that survives change or something?
• 0:29
  That depends on control. Welcome to AI for founders.
• 0:32
  I'm Ryan Estes. Today you'll learn how vision
• 0:34
  based autonomy is redefining robotics, why infrastructure is
• 0:38
  the silent killer of scale, and what deep tech founders
• 0:41
  understand about defensibility that SAS founders often miss.
• 0:45
  This episode is for builders who want moats, not marketing.
• 0:48
  If you want unfair advantages in AI every week, go to
• 0:51
  aiforfounders.co. It's where 27,000 founders
• 0:54
  sharpen their edge before their competitors wake up.
• 0:57
  It's one e-mail once a week and lots of big ideas.
• 1:00
  And if this made you think at all, leave a review like you're
• 1:03
  programming the algorithm to pay you back in karma.
• 1:06
  This episode is brought to you by Kick caster.com.
• 1:09
  If you're a founder who should be on podcasts but aren't, Kick
• 1:12
  Caster gets you booked on the right shows with the right
• 1:14
  audiences. Stop pitching yourself.
• 1:17
  Start building authority. Go to kickcaster.com and let
• 1:20
  your voice scale. That's Kick Caster by MO Burst.
• 1:30
  All right, March in. Welcome to the show.
• 1:32
  Thank you so much for joining me.
• 1:34
  Yeah. Thanks for having me, Ryan.
• 1:36
  Pleasure to be here. It's a pleasure to meet you.
• 1:39
  I'm really pleased that we are introduced, particularly because
• 1:44
  you know, it's rare that I get to work with somebody or have a
• 1:47
  conversation with somebody in robotics, particularly on the
• 1:49
  cutting edge of kind of visual automated robots.
• 1:54
  But equally kind of novel that you're also kind of the the
• 1:57
  developer from a hardware and software perspective.
• 2:02
  So let's dive right in. Your company is called seven
• 2:04
  sense dot AI. Give us kind of an overview of
• 2:09
  what you're working on. Yeah, of course.
• 2:11
  So, so basically we often say we built the eyes and brains for
• 2:15
  mobile robots. So our intention of the company
• 2:18
  was not to build the entire robotic systems, but actually do
• 2:22
  what we were really good at. So built both sensing, so
• 2:26
  cameras, different types of sensors and compute.
• 2:28
  So basically reasoning out of the sense data about like the
• 2:34
  environments to guide the mobile robots in an efficient way.
• 2:37
  So, so we felt there was a large, large gap in the market,
• 2:40
  especially when we started, right.
• 2:41
  So that was like 2018-2019 where industrial robots or mobile
• 2:46
  robots were like, we could say kind of dumb.
• 2:49
  So they were like mostly following some very well
• 2:52
  predefined path, very rigid, rigidly defined tasks.
• 2:57
  They could, they could barely adopt to changing conditions or
• 3:00
  to dynamic scenes. And that was a real challenge
• 3:03
  for for many applications. And we wanted to really change
• 3:07
  this right and bring a bit of intelligence to the equation.
• 3:12
  And I can say that the company was called 7th Sense Robotics.
• 3:16
  But meanwhile, in December 23, it was acquired by ABB.
• 3:19
  So, so I think that's also a very nice, you know, sign of
• 3:24
  actual value behind what we were doing, right?
• 3:26
  Because in the end, one of the largest robotics company or I
• 3:29
  think headcount was the largest robotics company, actually
• 3:33
  decided to internalize this technology, right for different
• 3:36
  kinds of robots. Amazing.
• 3:40
  Now, so you say you're making the eyes and the brain for for
• 3:45
  robots. And what I think is really cool
• 3:48
  from a fan perspective is how it looks.
• 3:52
  So if are you familiar with the movie Short Circuit?
• 3:56
  Unfortunately not. Oh my gosh, March in you you
• 4:01
  have a homework assignment. You must see Short circuit.
• 4:04
  It's possibly the best robot movie ever made.
• 4:08
  It's it's from the. 80s No, no. OK, then.
• 4:12
  I definitely need to catch up. Yeah.
• 4:14
  Yeah, so, so it, it was made in the 80s when I was a kid, and I
• 4:18
  loved it. It was about this, this, this
• 4:20
  friendly robot. But boy, if you made him mad,
• 4:23
  you, you look out, but the, the robot's name was Johnny 5.
• 4:28
  And your hardware looks exactly like Johnny 5's head.
• 4:34
  It's like it's almost identical. It's wild.
• 4:37
  Google it right now while we're talking, if you got a tab open
• 4:39
  Johnny 5, Johnny 5 from Short Circuit, I'm telling you, you're
• 4:43
  going to love this movie. OK, I see that now.
• 4:46
  Yes. Anyway, what's interesting is,
• 4:50
  is the application. So right now, just judging by
• 4:54
  your website, it looks like you're working largely with kind
• 4:56
  of like operators in warehouses. It looks like forklifts that can
• 5:00
  basically be autonomously manned.
• 5:04
  So this literally is kind of perched on top and becomes the
• 5:07
  eyes and the brain. So talk a little bit about the
• 5:09
  application so far that you've seen that have been successful
• 5:13
  with $0.07. Yeah.
• 5:14
  So, so I can say that let's say what you see on the website is
• 5:18
  one of the directions that we are following up more like
• 5:21
  commercially. So this is about indeed giving
• 5:24
  also this intelligence to man's forklift or industrial trucks
• 5:28
  because we've at some point realised that the same, the very
• 5:31
  same technology that we are deploying on autonomous robots
• 5:34
  can also work very, very well for forklifts which which are
• 5:38
  lacking a lot of this spatial awareness, right.
• 5:42
  So, so forklifts are normally not really aware where they are
• 5:46
  located. They, you don't really optimise
• 5:49
  the way how forklifts move. And also it's kind of funny,
• 5:53
  right, that forklifts are actually sometimes lost in
• 5:55
  warehouses. So sometimes they are parked for
• 5:57
  a break. Someone like goes away and when
• 5:59
  you have like 50 aisles and it's really, really hard where it's
• 6:02
  located actually. But let's say the majority or
• 6:05
  the most of the time and, and we are still doing this until this
• 6:10
  date, but with an ABB is basically equipping also
• 6:12
  autonomous mobile robots with the same technology, right.
• 6:15
  So it's, it's, it's basically a set of cameras depending on the
• 6:18
  on the on the application, it can be a few, it can be up to 8.
• 6:23
  And this, this brain, so basically an AI capable compute,
• 6:27
  which we all in a way, install a bit as a Tier 1 supplier to the
• 6:31
  robot and take control of the, of the vehicle, right?
• 6:35
  So, so you can maybe see that as a bit like, I don't know,
• 6:38
  companies that are or were doing autonomous driving, right, like
• 6:43
  Zooks or Lyft or companies that actually focus on providing a
• 6:49
  sensor suite and algorithms to actually drive the cars.
• 6:51
  So we are doing that for mobile robots.
• 6:54
  Amazing. So and I think I know a part of
• 6:57
  the novelty and potentially the IP is actually it's visual
• 7:01
  based. So give us kind of a
• 7:03
  differentiation between what it means to be kind of a, a visual
• 7:06
  base and, and kind of what the other options are.
• 7:11
  This episode is brought to you by a dangerous idea.
• 7:14
  It's called reading, specifically the AI for founders
• 7:17
  newsletter. 47,000 founders strong now.
• 7:20
  And it's the good stuff. No top ten AI tools that will
• 7:24
  change your life written by someone who doesn't ship
• 7:26
  product. It's just real breakdowns of
• 7:28
  what's actually working in AI, who's raised, who failed, who
• 7:32
  pivoted, and which founders quietly built $1,000,000 in ARR
• 7:36
  while everyone else was arguing on X.
• 7:38
  It's written for people building with AI every week.
• 7:41
  You get tactical frameworks, unfair advantages, breakdowns of
• 7:44
  real companies, and the occasional existential question
• 7:47
  about whether your startup is intelligent or just, well,
• 7:50
  branded. It's free, It's sharp.
• 7:52
  It might mildly offend your current workflow.
• 7:54
  Go to aiforfounders.co and subscribe before your competitor
• 7:57
  does, because the only thing worse than being replaced by AI
• 8:01
  is being replaced by a founder who reads.
• 8:03
  That's the newsletter at AI for founders.co.
• 8:06
  Now back to the show. Or yeah, so, so I can say that
• 8:09
  visual is very, very close to my heart.
• 8:11
  So somehow from from early on during my study.
• 8:14
  So I was like really excited about using cameras, especially
• 8:17
  like in embedded systems. So, So what we do is basically
• 8:21
  very, very close to what US humans use to perceive the
• 8:25
  environment. So, so we've used different
• 8:27
  types of signals. So we use cameras, right?
• 8:29
  Can be two, can be many, as I mentioned.
• 8:31
  So that's of course similar to our eyes.
• 8:34
  We are actually also using IMU signals and we as humans have an
• 8:39
  inner ear. So we have those like little
• 8:41
  stones in the ear that tears tell us where the gravity goes,
• 8:45
  right? And where, when do we stand
• 8:47
  upright And when we use also, if possible, odometry signals,
• 8:51
  right? So whenever you have wheels or
• 8:52
  legs or whatever, we can fuse this into our system.
• 8:56
  And you can imagine that our body does the same, right?
• 8:58
  We know that we are moving our legs and we know if we took a
• 9:02
  long step or a short step, other technologies.
• 9:05
  So usually that's lighter based, right?
• 9:07
  So you have two D lighters traditionally, so those that
• 9:10
  only scan a single plane, you have 3D lighters.
• 9:13
  So similar basically to to to autonomous vehicles, right?
• 9:16
  So way more or at that time Google was really Velodyne,
• 9:20
  right? Where like the leading companies
• 9:22
  in this initially 3D lighter technology.
• 9:24
  Now this is becoming way more common and then you have all
• 9:28
  kinds of things mounted on the floor.
• 9:30
  So starting with like black like optical recognition of a line or
• 9:34
  magnetic lines or maybe some inductive wires or QR codes
• 9:39
  placed on the on the floor. But the big difference for
• 9:41
  visual technologies of course that we can navigate using
• 9:45
  so-called natural features. So we don't need anything in the
• 9:49
  environment, but we just like recognize whatever is out there
• 9:52
  and use this to navigate, right. So it's very, very similar to
• 9:56
  you coming to a new place building in your brain a bit of
• 9:59
  a picture and map of this place and then localizing against
• 10:03
  maybe objects that you thought, OK, that's very specific to this
• 10:06
  place, right? So I'm going to remember it now.
• 10:09
  Yeah. Well, in to to have kind of
• 10:11
  vision base, it's probably presupposed that you have to
• 10:14
  have a pretty high level of processing and you have to have
• 10:18
  a pretty high level of compute like it wants to be smart.
• 10:21
  Otherwise there's just so many inputs if you're actually
• 10:24
  visually mapping as opposed to using a LIDAR based kind of flat
• 10:27
  dimensional kind of plane. Now, I would love to hear kind
• 10:32
  of your take as a designer for the product, because when I'm
• 10:36
  looking at it, it looks like Johnny 5 and which is to say
• 10:41
  that it has eyes similar to us where it's two eyes facing
• 10:44
  forward. Now I'm a sportsman, OK?
• 10:47
  And so the way I kind of judge what I hunt for food is largely
• 10:53
  on the position of its eyes. You know, so if it has eyes up
• 10:57
  front, this is a predator. You know, typically your
• 10:59
  predation kind of animals or mammals at least have eyes
• 11:03
  facing forward. They're very good at, like,
• 11:04
  honing in on sharp details. And then you have what I look at
• 11:08
  as food that has eyes on the side, you know, whether it's
• 11:12
  like, you know, cattle or it's deer or elk or moose or these
• 11:17
  kind of animals where they're scanning the periphery, largely
• 11:20
  they can't see necessarily very sharp and focused, but they can
• 11:23
  see behind them. But then you go even a step
• 11:25
  further into the animal Kingdom, if we're thinking about eyes and
• 11:28
  you have like dragonflies and spiders and they have eight eyes
• 11:32
  and different, different ways that they're processing vision.
• 11:35
  So as a designer, how do you start to develop a product from
• 11:39
  the eyes perspective? That's, that's a great question
• 11:43
  And it's funny, right, that you raised this here and it's, it's,
• 11:46
  it's, it's a constant topic with all our customers.
• 11:48
  Where do we place cameras, How do we place cameras?
• 11:51
  And that's really application specific.
• 11:53
  So what you see on our website is a is a very simple set up, I
• 11:56
  think with free cameras, but often on, if I'm not mistaken,
• 12:01
  but often in our products, we are using 6 or even 8 cameras,
• 12:05
  placing them all around the robot, right?
• 12:06
  So we see this a bit as a, as a robustness measure.
• 12:11
  So often we work pretty well with cameras like or eyes,
• 12:15
  right, like facing one side, but it's always beneficial to see
• 12:20
  more. So whenever robots like is like
• 12:23
  the cameras from one side are completely concluded occluded,
• 12:26
  sorry, by a load or occluded by some structure in the
• 12:28
  environment or maybe they're just looking at the white wall.
• 12:31
  It's always nice to have sensors looking at actually something
• 12:35
  meaningful from the other side, right?
• 12:37
  And the cost is kind of limited. So, but yeah, it's a, it's a
• 12:41
  very good point, right about those two eyes more or less
• 12:44
  similarly spaced to humans. That's actually like coming from
• 12:48
  physics from geometry. So if you want to recognize
• 12:51
  objects around the robot somewhat close to the robot,
• 12:55
  right, within this range of 1 to 3 meters, this spacing of so, so
• 13:00
  this baseline so-called of, of cameras in this stereo
• 13:04
  configuration actually needs needs to be around this 10 to 12
• 13:07
  centimeters so or that's the optimal distance and that's
• 13:11
  where it comes from. So it's really nature or
• 13:13
  evolution kind of perfected this, right?
• 13:15
  Because that's the distance where we actually manipulate
• 13:18
  objects. Interesting.
• 13:20
  It's it, Yeah. It's, it's super fascinating
• 13:23
  Now. I feel like, you know, the, the
• 13:26
  everybody's looking at like the robots in China that are doing
• 13:29
  Kung Fu and like all these new kind of like videos that come
• 13:32
  out all the time of the, the advancements in kind of like
• 13:36
  these kind of humanoid type robots, androids, you know, and
• 13:42
  a lot of consideration has to go into like how, what's the
• 13:45
  personality kind of like, you know, so like when you watch
• 13:49
  short circuit, you'll see it has like these really cute, like
• 13:52
  kind of like flaps for eyebrows, which emote, which is because
• 13:56
  otherwise how can you sense its emotions without eye movement?
• 13:59
  There's a big reason why people like dogs and they don't like
• 14:03
  cats, because cats don't have the muscles in their face to
• 14:06
  show emotion like dogs do. Now you have an advantage
• 14:09
  because you're using industrial based robotics for very
• 14:13
  practical sense. It's not going to make you a
• 14:15
  cheese sandwich at least yet, you know.
• 14:17
  But how do you kind of approach the the, the design from just
• 14:21
  like a, a tactical integration into kind of human culture and
• 14:26
  like kind of what humans will accept as being kind of a
• 14:28
  companion? That's, that's a very good
• 14:31
  question. So I can say that one thing that
• 14:34
  goes a bit into this eyebrows topic, right, is that how do you
• 14:37
  communicate from the robots to humans around it, right?
• 14:41
  And like in industrial spaces, this is like you don't really
• 14:46
  have, you know, those fancy facial features, right?
• 14:48
  But you usually have some kind of lights, you have sound, maybe
• 14:51
  you have a display and you try to like make.
• 14:56
  I think the most important thing is about humans actually
• 14:58
  understanding what the robot is intending to do.
• 15:01
  So the classic, which is now by now mandated by safety
• 15:05
  standards, is actually blinking when when you want to turn right
• 15:08
  or changing the light colour when you're speeding up.
• 15:12
  So basically giving a bit of intention science to the robot
• 15:17
  is really, really helping the acceptance of robotics in our
• 15:21
  experience. And I think in the general
• 15:23
  market experience. The other side is actually
• 15:25
  recognizing humans, right, and reacting a bit differently among
• 15:29
  them compared to, you know, like a parcel or like a palette with
• 15:34
  some goods. So we did quite a bit of work,
• 15:38
  especially when we were working on cleaning machines.
• 15:40
  So those that operate in public spaces, like professional
• 15:43
  cleaning machines for airports, railway stations and so on, to
• 15:46
  recognize humans, try to predict their emotion and also react
• 15:51
  accordingly to to to very motion movements, right?
• 15:54
  So, so the robot was not just passively following the
• 15:57
  trajectory and stopping in front of an obstacle, whatever that
• 16:01
  was, but it was actually trying to kind of mingle between people
• 16:05
  in a way that would be at least disturbing for for passers by as
• 16:10
  possible. So there is a specific actually
• 16:12
  domain of robotics dealing with this social side of mobile
• 16:15
  robotics which I find quite interesting.
• 16:18
  Yeah, me too, you know, and that's kind of like where we're
• 16:21
  at. It feels like kind of tip of the
• 16:22
  spear as far as intelligence goes is the, the predictive
• 16:25
  nature that AI gives us that can inform kind of robotics
• 16:31
  movement. So on on one side, they, they
• 16:33
  need to signal and it's important that they're signaling
• 16:36
  their movements because they're moving in space and it's
• 16:38
  important for them to signal maybe the, the, the health, for
• 16:42
  lack of a better word, you know, So, so signaling is really
• 16:45
  important and then also analyzing its behavior,
• 16:49
  particularly around humans, predictive of what they're going
• 16:51
  to do. So they cannot occupy the same
• 16:53
  space as humans because ultimately nobody wants a robot
• 16:56
  in their way. What else kind of really go
• 17:00
  ahead please. Just wanted to say one more
• 17:01
  thing that came to my mind. I mean, one example that I find
• 17:04
  really, really interesting is how two groups of people or
• 17:08
  humans pass each other on a sidewalk, right?
• 17:11
  So if you're like walking towards a person coming from the
• 17:15
  opposite side on a sidewalk, where is a bit of this, you
• 17:18
  know, looking at each other negotiation, like how do you
• 17:21
  actually navigate exactly, exactly, right?
• 17:24
  And that's exactly what AI and robotics needs to do, right?
• 17:28
  It needs to figure out, will the human yield way?
• 17:31
  Should I yield way? Should I go left?
• 17:33
  Should I go right? And that's exactly this art of
• 17:36
  making it kind of friendly and acceptable towards humans, right
• 17:40
  when it's behaving kind of like a human.
• 17:45
  That's interesting. So what else goes into it?
• 17:48
  So there's signalling, there's conscientiousness, for lack of a
• 17:52
  better word, that it's going to like move out of the way.
• 17:55
  What else is kind of important when you're thinking about, of
• 17:57
  course safety is going to be a primary concern, What are some
• 18:01
  other concerns when we're thinking about the intelligence
• 18:03
  of the robots? Yeah, so, so from my
• 18:06
  perspective, so coming from Sevensence, I have to say like
• 18:09
  spatial awareness, right? So, so many times robotics or
• 18:12
  humanoids operate in some confined space, right?
• 18:15
  We kind of manipulate some objects, right?
• 18:17
  We see dance or whatever. But if you want robots to re
• 18:21
  navigate a city or navigate the entire building smoothly, they
• 18:24
  need to know very well in a way where they are located and what
• 18:28
  are the options to get from one place to the other, right?
• 18:30
  And that's something that is very, very close to our hearts
• 18:32
  at Sevensence. I think the other part, which is
• 18:36
  I think only solved to a limited extent, right?
• 18:39
  It's really understanding physical interactions in the
• 18:42
  environment, right? And that's where cameras also
• 18:44
  come very useful because you can recognise objects, you can
• 18:47
  recognise how they are like located with respect to each
• 18:50
  other, right? So, so, so often in case of
• 18:53
  humanoids or some manipulators, you really want to have a good
• 18:56
  picture of two objects and understand from the AI
• 18:59
  perspective, right, how you can manipulate them.
• 19:03
  Are they stick together? Are they separate?
• 19:05
  Can you connect them? Can you disconnect them?
• 19:08
  All of those interactions with objects.
• 19:12
  Again, thinking about this industrial topic or the topic of
• 19:15
  like, for example, medical facilities where we would like
• 19:20
  mobile robotics to one day help automate the labs, for example,
• 19:25
  right? You would like them to be as
• 19:27
  smart as possible and be able to manipulate objects in a in a
• 19:31
  nimble but also smart way where you don't need to define every
• 19:35
  millimetre or inch of movement, right?
• 19:38
  Yeah, It can interpret it and figure it out kind of like a,
• 19:42
  like a cat. Like if you move a chair in a
• 19:44
  room and you see a cat come in and it like, looks around for
• 19:47
  like a second and it's just like something's different, you know?
• 19:50
  And then it's got it. Yeah.
• 19:53
  Yeah, or, or, you know, you could also like think of like
• 19:58
  really babies, right. Learning how to grasp and
• 20:00
  actually somehow learning how to touch objects and and what you
• 20:04
  can do with them, right. So our kids or our kids had this
• 20:08
  face where he was like shaking every object you could find when
• 20:11
  he was, I don't know, whatever, a few months old maybe, right,
• 20:14
  Because some of the objects like rattled, right.
• 20:16
  And he was like testing every objects for doesn't rattle.
• 20:19
  And then there was this. This, this face of stretching,
• 20:22
  right? Because he recognized some
• 20:23
  objects are like silicone or rubber and he can like stretch
• 20:26
  them so every or or maybe tear them.
• 20:28
  So basically he was picking up random objects and trying to see
• 20:32
  what happens if you kind of pull, oops, sorry, pull them
• 20:34
  apart, right? And that's that intuition is
• 20:39
  built in humans when we are a one year old, but we would like
• 20:42
  the same for robots to know, right?
• 20:45
  You describe that as curiosity. You want the robots to be
• 20:48
  curious a little bit. There's one way of learning it,
• 20:51
  right? So you can somehow embed this as
• 20:54
  a package of prior knowledge about the world, right?
• 20:58
  The other way is basically let them experiment the same way as,
• 21:02
  you know, kids and babies do, right?
• 21:05
  Of course there are like two ways.
• 21:08
  Interesting. Yeah, Now, you know, it seems
• 21:10
  very smart to to be in business in robots as with in like
• 21:14
  logistics and manufacturing, largely because at work we're
• 21:17
  we've been comfortable with robots for a long, long time.
• 21:21
  As the it's starting to move into the home.
• 21:24
  What is your take? And I'll put into context like
• 21:26
  this. You know, Elon last week said,
• 21:28
  you know, we're not going to need doctors because the robot
• 21:31
  that lives in your house is going to be the best surgeon the
• 21:33
  world's ever seen. So if you need a medical
• 21:35
  procedure, like your robot will just do it for you.
• 21:39
  Now, I am not so convinced that people are going to want robots
• 21:42
  in their home. You know, if you look at kind of
• 21:45
  the history of failed products you have like the Google Glass,
• 21:48
  you have the Apple Vision Pro, you know, there's certain things
• 21:52
  that people are just like. And so it, it feels like robots
• 21:57
  coming into the house, it's going to have some iterations.
• 22:00
  And I'm just not convinced. The first first iteration is
• 22:03
  going to be great, largely because, you know, the
• 22:06
  autonomous robots that they've placed in the home so far are
• 22:10
  actually not autonomous. They're being piloted by some
• 22:12
  guy in the Philippines. Yeah, Yeah, exactly.
• 22:15
  Yeah. So what what's your opinion on
• 22:18
  kind of robots moving into the home and is that kind of a
• 22:21
  frontier you kind of want to to pioneer as well?
• 22:26
  Yeah. So, so maybe returning to this
• 22:28
  industrial side first, right? So, so at some point in the in
• 22:32
  the history of seven cents, right, we realized that we need
• 22:35
  to focus on the market that actually exists, right?
• 22:37
  So especially in like being a Tier 1 supplier, we cannot wait
• 22:42
  for robot builders to appear 10 years later, right?
• 22:45
  If you have a company and need to fundraise and you need some
• 22:47
  revenue at least. So, so industry is definitely
• 22:51
  the, the, the place where robots are accepted, right?
• 22:54
  I think what is changing and over the course of last whatever
• 22:58
  10-7 years, right is that we are more and more used to robots on
• 23:02
  sidewalks, right? Maybe some half autonomous cars
• 23:05
  or autonomous cars. We are getting used to robots,
• 23:09
  like cleaning robots in public spaces, maybe delivering robots.
• 23:12
  But I, I see your point, right? It's really hard to cross this
• 23:15
  barrier of someone's privacy, right?
• 23:17
  And someone's home, right? Maybe we're OK with a, with a,
• 23:21
  with a robotic vacuum cleaner, right?
• 23:22
  Or like a mopping robot right now at home, but having this
• 23:27
  personal assistant walking behind you at home, it's like a,
• 23:31
  it's a, it's a big step in terms of trust towards like data usage
• 23:35
  and privacy. I, I would say that's the
• 23:36
  biggest problem, right? Of course, right now building a
• 23:39
  humanoid just for everyone's private use is also like kind of
• 23:44
  an expensive it, it's hard to sell right in on mass in
• 23:47
  1,000,000, but I'm sure this is going to change, right?
• 23:49
  It's when we were starting or when I was starting to play with
• 23:53
  embedded vision, right? It was like completely
• 23:54
  unthinkable to have embedded vision in a consumer product.
• 23:58
  Like really popular, right? By now you can basically get 100
• 24:02
  bucks drone that has a camera and does some basic whatever
• 24:06
  human trucking, right? Yeah, it's crazy.
• 24:10
  It, it's crazy, you know, and I, I think it largely has to do
• 24:15
  with the emotion behind the form.
• 24:20
  Like, which is to say, I think if these robots in the robot
• 24:23
  army comes out and they look like dogs or they have a dog
• 24:26
  form, like people would be much more willing to like have them
• 24:29
  in their house because it because people love dogs.
• 24:34
  But if it looks like a little person, like a four foot human
• 24:37
  robot. Now you, you were talking about
• 24:40
  like that dance you do when you're walking towards somebody
• 24:42
  and you're, you're trying to figure out like, am I going left
• 24:45
  or right or whatever? You know that that is kind of
• 24:49
  like alludes to the fact of like why, you know, in the States, we
• 24:52
  drive on the right hand side, you know, and they go on the
• 24:56
  left and why in Britain they it it's reversed and for Europe
• 24:59
  it's it's mostly reversed. And that's largely because, you
• 25:03
  know, most people are right-handed, which is to say
• 25:05
  when they were jousting on, you know, horseback and they had
• 25:09
  lance's, you would be on the left side of your adversary.
• 25:13
  You know what I'm saying? Or the reason why we, the reason
• 25:16
  why we shake hands when we greet each other is because we're
• 25:19
  demonstrating that we don't have weapons.
• 25:21
  You know, humans, it's, we're naturally violent creatures.
• 25:26
  And like we're, we're looking at other humans skeptically.
• 25:29
  And so like, perhaps it's Hollywood, but we look at that
• 25:32
  thing almost adversarially potentially.
• 25:35
  So it, it seems like to me, it's like if you're going to make
• 25:38
  something that is, you know, has two arms, 2 feet, why don't you
• 25:42
  just have it walk on all fours, give it a dog face and then
• 25:46
  suddenly everyone's gonna love it right out of the gate.
• 25:48
  Because I would feel much more comfortable if I had a little
• 25:52
  doghouse where my robot lived and I could like put blankets on
• 25:55
  it and take it for a walk, you know, then if it's just some
• 25:59
  creepy thing standing in the, in the my closet waiting for
• 26:02
  commands, you know what, what do you think about that?
• 26:05
  How do you feel about the emotional reception of robots?
• 26:09
  So I mean, I, I agree with that, right?
• 26:11
  So, so where is this kind of uncanny valley topic, right?
• 26:14
  Where a bit too close to a human being, we're feeling kind of
• 26:18
  awkward, right? So as long as it's a very
• 26:21
  clearly a robotic device, that's fine, right.
• 26:25
  But when it's getting close to a human, it is a bit creepy.
• 26:28
  It is a bit scary, it makes us uncomfortable.
• 26:31
  What challenge that I see in this, right is that if we want
• 26:34
  robots in our homes, it's in the industry, we would call it a
• 26:37
  brownfield facility, right? So basically something that is
• 26:39
  already built, created and equipped.
• 26:42
  And if you want to manipulate things that we manipulated our
• 26:46
  homes, you probably need to be more or less of a similarly tall
• 26:52
  as we are, right, and have hands more or less at the same height
• 26:55
  as we do. And that creates a bit of a
• 26:58
  problem, right? For a dog.
• 26:59
  I mean, it's a funny robotic pet, but if you like.
• 27:03
  I mean, I would love if, if, if, if a robot, a humanoid could
• 27:07
  fold my laundry, right? Or like, you know, do some
• 27:09
  household chores. But in the end, you probably
• 27:12
  need to be somewhat similar to a human with two manipulator arms,
• 27:16
  right? Otherwise it's going to be
• 27:17
  really hard. So yeah, it's a challenge.
• 27:22
  And we'll see. I mean, this is going to come
• 27:24
  very quickly, you know, so we'll see.
• 27:26
  I mean, and I'm, I'm saying that now, but I'm probably signing up
• 27:29
  for that first Elon robot. You know, I might have to get
• 27:32
  one of those and just see, it's definitely interesting, right?
• 27:36
  Like exploring the the the limit of of of the technology and you
• 27:41
  know, robotics at the time when it ships, right.
• 27:43
  That's going to be for sure exciting, right.
• 27:45
  Yeah, totally. I.
• 27:47
  Want to learn a little bit more about you?
• 27:48
  March in and I want to give a shout out to Kevin.
• 27:52
  Fan of the show, Friend of the show, interviewed him.
• 27:55
  Possibly the best podcast app in the world, which is snip.com.
• 28:01
  It's incredible. And, and I really am grateful
• 28:03
  for Kevin for introducing us. And you know, when I was talking
• 28:06
  to him, I was very curious about like what's happening in
• 28:09
  Switzerland right now, because Switzerland seems to be maybe
• 28:12
  tip of the spear as far as like everything cutting edge in
• 28:16
  technology, which is AI and robotics straight up.
• 28:19
  And you share that capacity. So maybe I'd love to hear a
• 28:23
  little bit of your story of like how you came into this field and
• 28:27
  how you got started. And then also kind of what's the
• 28:29
  culture in Zurich, Switzerland right now?
• 28:34
  Yes, so so. I can maybe start from my story
• 28:37
  and then this kind of smoothly leads us towards Switzerland.
• 28:40
  So, so I originally come from Poland, from a city called
• 28:43
  Frotswaf and I studied there for, I think for years.
• 28:47
  And at some point I, I, I just wanted to, you know, see some
• 28:51
  other place, study at some other place, right?
• 28:53
  Just take, take a, take a bet and see how it goes.
• 28:57
  And it was pretty late to start the following year, and I mean,
• 29:02
  quite randomly, it happened that the only university still
• 29:05
  accepting a kind of second rounds of recruitment were in
• 29:08
  Switzerland and I think in, in the Netherlands.
• 29:11
  So I applied to to Turek and also Lausanne and and got
• 29:16
  accepted and and basically moved to Zurich for the other part of
• 29:19
  my studies, not, not really thinking, I must admit about the
• 29:24
  country at that time, right? So I was really driven by kind
• 29:27
  of engineering interests and wanting to kind of see a really,
• 29:30
  really great university that this ETH Zurich.
• 29:35
  So this Technical University here is and I, I did study
• 29:40
  system like electrical engineering and systems and
• 29:43
  control, but in the end ended up in a lot of robotics courses and
• 29:46
  very quickly realise how amazing this robotics community here is.
• 29:51
  So we're like already at that time a bunch of very good
• 29:54
  robotics lab and great professors.
• 29:57
  So big shout out to Roland Ziegwart, one of the, you know,
• 30:00
  founding fathers of this whole robotics and AI craze here.
• 30:05
  And, and I got like absolutely in love with what was happening
• 30:09
  in the lab, right? So I did my projects there as a
• 30:11
  student when I kind of smoothly transitioned into the PhD doing
• 30:16
  all kinds of like visual sensing, mapping, localization,
• 30:19
  place recognition. And it happens that at more or
• 30:23
  less this time the community around AI or BAI, but visual
• 30:28
  sensing was really growing fast in Switzerland.
• 30:30
  So there was already an office of Google here.
• 30:34
  So at that time that was around probably over 1000 people
• 30:37
  actually. But also Apple started Meta or
• 30:42
  at that time Facebook with ARVR technology started, Microsoft
• 30:47
  starting the HoloLens project here, and also a bunch of
• 30:51
  others. And of course there were like a
• 30:54
  few robotics spin offs from ETH, maybe one of them.
• 30:58
  Interesting one is antibiotics, right, with those quadruped docs
• 31:03
  which we're also growing. So in the end, there is this
• 31:06
  like really vibrant community of roboticists and AI engineers
• 31:09
  here. And I think this is fantastic,
• 31:13
  right? Because if you look at
• 31:14
  Switzerland and the fact that the labour is very, very
• 31:18
  expensive here, right? It's maybe comparable to Bay
• 31:20
  Area, but it's like way more expensive than any other place
• 31:24
  maybe apart from Norway in Europe.
• 31:25
  So the only chance to actually find job for those people is
• 31:31
  basically having something which is high value at right.
• 31:33
  And basically building robotics technology or AI is actually
• 31:37
  what's what's, what can keep the market really going and and keep
• 31:40
  the whole country successful. And that's the story of
• 31:44
  Switzerland, basically, right? Always do something which is
• 31:47
  cutting edge that others cannot do equally well.
• 31:50
  Incredible. So you you achieve your PhD, you
• 31:56
  earn your PhD, you have an idea for a company.
• 32:00
  How did you go about financing the company?
• 32:03
  So. Like what is very, very
• 32:06
  convenient is also at least for for for European standards.
• 32:10
  I'm not sure about the US, but Etihad.
• 32:13
  So ETH, this university is very, very friendly to its founders.
• 32:16
  So it's relatively easy to spin off and transfer on like pretty
• 32:21
  good conditions. Your technology coming from from
• 32:24
  your university. Stay into the company.
• 32:28
  So that's what we did with a group of founders and we
• 32:31
  actually stayed in some kind of incubator for a while within the
• 32:35
  the university, which provided some seed founding and also, you
• 32:39
  know, basics like space, also access to workshops of the
• 32:43
  university, which is pretty important, right?
• 32:45
  If you're like a hardware company, you know, some basic
• 32:49
  infrastructure, let's call it like this.
• 32:52
  I think the initial, the initial financing was also coming from
• 32:55
  the network of investors in Switzerland.
• 32:58
  So we're like a few Angel investors, a few big companies
• 33:02
  actually very interested in robotics.
• 33:04
  And how did this is how it kicked off And for our Series A
• 33:08
  rounds, right? We, we already had some global
• 33:11
  reach and we could, we could reach basically globally for
• 33:15
  investors, right? So we had a bit more opportunity
• 33:19
  to to fundraise, but the first step was really like the local
• 33:22
  market. And they're a somewhat robotics
• 33:24
  community of Switzerland and maybe the countries like
• 33:28
  Germany, right? So neighboring countries that
• 33:31
  invest heavily in in, in Swiss startups.
• 33:34
  Yeah. So you.
• 33:35
  You. Demonstrate proof of concept and
• 33:38
  then you go get your Series A. Yes.
• 33:41
  Yes, I would say so. So, so we were very, very lucky
• 33:44
  because our first big project was about this autonomous
• 33:47
  cleaning machine that I mentioned before.
• 33:49
  So professional cleaning machine, scrubber dryer machine
• 33:51
  for airports and so on. Commercial sensors.
• 33:55
  I'm sorry, flooring. Yeah, it's like.
• 33:59
  Yeah, cool. And.
• 34:02
  That was basically a local Swiss company, relatively big market
• 34:06
  player in Europe. That was interesting in building
• 34:09
  a smart autonomous machine. And you know, they contacted the
• 34:12
  university contact university, really forward them to us and
• 34:16
  we've built it together, right? So, so it was really, really
• 34:20
  tough, I can say, right, because before you built like all the
• 34:23
  proper delivery pipeline of software and build your hardware
• 34:27
  and you're like a mature start up, we had to deliver products
• 34:30
  that actually work in the hands of non roboticists.
• 34:34
  But in the end, I think that was, you know, it was also like
• 34:39
  an experience that probably gave us a gigantic opportunity,
• 34:43
  right? So basically having revenue,
• 34:45
  having a product in the market and also having a robust
• 34:48
  technology very early on for start up standards.
• 34:51
  And that's probably what were sales, what were sales like?
• 34:54
  For the the robot cleaner. So I cannot really comment.
• 34:59
  On the numbers, it's still being sold, but I can say it's like
• 35:03
  it's basically like few 100 units that are being used.
• 35:07
  Amazing how much. Is each unit.
• 35:11
  So again, cannot comment on. That, but I can say that the
• 35:14
  usual pricing is of of the end product, right.
• 35:17
  It's around 20 to 50K per unit I think.
• 35:20
  Wow, amazing. What an incredible win out of
• 35:23
  the gate. So you you develop a commercial
• 35:25
  project before you really even have the start up together.
• 35:28
  And so based on the success of that, they're like, what do you
• 35:31
  want to do? Here we go.
• 35:33
  Yeah, that's all. Yes, exactly.
• 35:34
  And you know when, when we enter this industrial stage, right?
• 35:38
  So I'm talking about mobile robots for industry.
• 35:41
  Like my perception was that at that time we were extremely
• 35:44
  robust compared to the kind of situation in the market.
• 35:48
  So we were able to, you know, after kind of fixing all the
• 35:50
  bags and really tough times, right, with the cleaning
• 35:53
  machine, right? And I let's be honest, it's not
• 35:56
  always, it wasn't always pretty, right?
• 35:57
  So a lot of times we had to, you know, stay long the hours or
• 36:02
  weekends and try to fix it because customers were like,
• 36:04
  hey, that doesn't work at all, right?
• 36:07
  And but in the end, after after fixing all of that, that's after
• 36:12
  going through this miserable times in the from time to time,
• 36:17
  in the end, you end up with a like really mature system,
• 36:19
  right? That's that you can go to an
• 36:21
  industrial partner to manufacturing and say, Hey,
• 36:23
  look, this, this, this technology will actually like
• 36:26
  work for you 24 by 7, like free shifts, right?
• 36:29
  And I think that's something which is really, really
• 36:32
  difficult without this flagship project, right?
• 36:34
  Or this project that guides you initially as a start up.
• 36:41
  Incredible, can you comment? On how much you raised for your
• 36:44
  series. AI think that's.
• 36:48
  Public. So I think that was I think in
• 36:51
  the USD 7.7 million, I believe if I remember correctly.
• 36:55
  It's been a while ago, but I I think so now, yeah.
• 36:58
  And then how do you feel? About runway now, are you guys
• 37:00
  profitable? How does company look now?
• 37:03
  Yeah, so, so as I've mentioned, right.
• 37:05
  So the company was actually fully acquired by ABB at the end
• 37:09
  of 23. So, so we are basically a part
• 37:11
  of, of, of ABB Robotics. So now it's getting interesting
• 37:15
  because ABB Robotics is carved out from ABB and got acquired by
• 37:20
  SoftBank. So it's going to be a kind of
• 37:24
  another new start for us, right, Starting more or less next year.
• 37:28
  Interesting. So you'll have a new.
• 37:29
  Job or a new job title? Same job, Yeah.
• 37:33
  Yeah, exactly. New employer, same title, I
• 37:36
  guess. Yeah, I I wonder, do you see
• 37:38
  that? Happening in Europe that kind of
• 37:40
  like, you know, the trend for a while was private equity was
• 37:43
  rolling up companies and now it seems like they're starting to
• 37:46
  separate them. So carving out ABV robotics and
• 37:49
  and selling that unit to another.
• 37:51
  Is that kind of happening more and more in the startup scene
• 37:53
  out there? That's.
• 37:58
  I'm not sure if I see that right.
• 37:59
  So, so ABB is a bit of a special case and I can say that, you
• 38:04
  know, there are like of course, many fears when your start up is
• 38:06
  getting acquired by, you know, hundreds years old industrial
• 38:11
  group, right. But I think ABB is is really
• 38:15
  special, right. So that's a company that somehow
• 38:19
  really proved itself over time and managed to stay agile and
• 38:23
  fairly like fairly flexible for like an organization of this
• 38:27
  size in all the market changes. And in case of ABB, there is
• 38:32
  more like that focus on specific areas driving what is being kept
• 38:39
  or carved out. So the primary business for ABB
• 38:43
  is basically motion like electric drives and
• 38:45
  electrification. So all the power plants and
• 38:49
  electric infrastructure products, while robotics was a
• 38:52
  bit of a also size wise, right, a bit of a site, site
• 38:58
  organization. And at some point I think there
• 39:01
  was a realisation that it's probably worth to to sell it and
• 39:03
  really focus on the core business.
• 39:05
  Also robotics as we know is, is quite a bit different, right,
• 39:08
  when building power generators, right?
• 39:11
  It's a lot about AI, it's a lot about software.
• 39:14
  So I think it's a smart decision, right?
• 39:16
  That like the priorities and the way how you operate actually
• 39:20
  differs between, you know, building robots and building
• 39:25
  electrical or electrification or motion products.
• 39:29
  So, you know, I'm getting carved out, but I feel it's, it's a
• 39:34
  really sensible decision, right? And it's it's, it's a pattern
• 39:37
  that ABB does with a lot of businesses and somehow it
• 39:41
  guarantees that it stays focused and doesn't grow into just too
• 39:45
  many areas in too many fields, 100%.
• 39:49
  Amazing. All right, March in.
• 39:50
  We're about to enter the Quantum round.
• 39:52
  Are you ready for this? Yes.
• 39:57
  Cool. We'll start with the founder
• 39:58
  Challenge question. This is brought to you by
• 40:00
  kitcaster.com. OK, what's the most money you've
• 40:03
  wasted on something completely useless?
• 40:09
  The founder. Good question, I mean.
• 40:14
  One thing that comes to my mind. Is that being a product person
• 40:17
  right now myself, I think initially we were so tech
• 40:20
  driven. We were like happily burning
• 40:22
  money on developing technology, which we thought is really cool,
• 40:27
  but was not really driven by any market need.
• 40:30
  So it's not really like maybe question is about a thing a
• 40:33
  product that we bought, but it's actually burning money on R&D
• 40:36
  nobody needs, but it's exciting was really a trap for us, you
• 40:39
  know, coming from engineers coming from from the university
• 40:44
  basically. But that's important part of the
• 40:47
  the. Process though, right?
• 40:48
  It's kind of like you could be like, what's the point of
• 40:50
  dancing? It's bad.
• 40:51
  It's not good cardio. You're not getting stronger, but
• 40:53
  you're like, no, you if you dance.
• 40:56
  Yeah, yeah, yeah. So yeah, it, it was a very
• 40:59
  interesting experience and a bit of a, you know, like a aha
• 41:04
  moment in a way, right? You bet.
• 41:07
  OK. Your company's name is $0.07.
• 41:09
  So sight, hearing, taste, smell, touch, What are the other two
• 41:14
  senses? Nice.
• 41:17
  Nice try. That's a question that appears,
• 41:19
  but we saw it as six degree of freedom estimation in time
• 41:22
  sometimes. So XYZ, then the rotations, so
• 41:26
  pitch, roll, pitch, roll, Yaw, and then time.
• 41:30
  Time I love. That interesting?
• 41:33
  OK, cool. Now what I want to do is I want
• 41:38
  to see if we can ascertain your favorite robots from pop
• 41:42
  culture. Now, obviously this is going to
• 41:45
  be a lot of American movies, so if you don't know what it is,
• 41:48
  you can say skip, but I'm going to give you a this or that robot
• 41:52
  to go versus each other. And I want to hear what your
• 41:54
  favorites are. OK, you ready?
• 41:56
  OK, let's let's give it a shot. Yeah, I'm going to start with
• 41:59
  how. From the Space Odyssey versus
• 42:03
  R2D2 R2. D2 definitely.
• 42:07
  R2D2 Verse. C3PO Oh, that's a tough one.
• 42:13
  I think I'm going to stay with R2D2.
• 42:15
  That was my favorite when I was a kid.
• 42:17
  He's the best, right? I noticed.
• 42:19
  Like the the your visual visual placement robot looks a little
• 42:24
  bit like R2D2, I'm not going to lie.
• 42:27
  OK R2D2 or BB? 8 I think.
• 42:33
  I think I'm going to stay with R2D2.
• 42:35
  I'm sorry you kind of feel quickly got my favorite right?
• 42:39
  It's a kids favorite so hard to argue with this one.
• 42:42
  OK. R2D2 versus T8?
• 42:44
  100 from the Terminator. Yeah, don't want to stick to
• 42:52
  R2D2, but I mean, the reality is it's really the cuteness of R2D2
• 42:56
  thinking about our discussion from the past, right?
• 42:58
  It's, it's really this thing that you could imagine having at
• 43:01
  home and it's kind of cute, you know?
• 43:03
  And he kind of whistles. At you, which is great way to
• 43:06
  signal, you know, kind of music and melodic, kind of like a
• 43:09
  bird. I love that R2D2 verse Optimus
• 43:13
  Prime. Maybe.
• 43:18
  Maybe. Optimus Prime is actually a fair
• 43:21
  competition to R2D2. It's a pretty good one.
• 43:23
  Let's go with that. Optimus Prime verse RoboCop
• 43:28
  Optimus Prime Sorry. Optimus Prime versus.
• 43:32
  Data from Star Trek. Yeah, that's also brings child.
• 43:38
  Memories, to be honest. So, you know, when I was a kid,
• 43:42
  you could watch Star Trek in Polish television and.
• 43:46
  And I was really exciting to, to to see that, right.
• 43:49
  And I was like, really, really exciting about Data.
• 43:53
  Yeah. Maybe Data.
• 43:55
  OK Data. Data was really cool too,
• 43:57
  because he he kind of does represent where we're maybe that
• 44:01
  this time we're at right now, where the it's it's really good.
• 44:05
  AI is really good, but also it's it's learning.
• 44:07
  It's like it's not quite there. There's a there's a brilliance,
• 44:10
  but there's also a very childish innocence to it a little bit.
• 44:14
  OK, data verse Wall-e The Pixar movie.
• 44:22
  Yeah, I was when you when you mentioned.
• 44:23
  Like pop culture, almost my first May 1st of all went into
• 44:26
  Wall-e. So it is cute, right And and it
• 44:30
  kind of Pixar. Pixar is always like making a
• 44:34
  great job, but making characters lovable, right.
• 44:37
  So totally agree, I would say, OK.
• 44:40
  Wally versus Eve. I think Wally, yeah, yeah.
• 44:46
  OK, how about Wally versus the Iron Giant?
• 44:54
  I'm going to stay with Wally. OK, Wall-e.
• 44:58
  Verse R2D2. That's a that's a tough one,
• 45:04
  right. So, so as I said, like where is
• 45:06
  a bunch of those robots where where you kind of built your
• 45:09
  emotions? So it's right.
• 45:11
  And R2D2 and wall-e are probably like really it's a it's a draw
• 45:15
  in a way for me. Yeah, yeah.
• 45:18
  A draw and we'll end with the draw.
• 45:20
  You know, my favorite R2D2 scene is Return of the Jedi when
• 45:25
  they're about to March Luke off the the walk the plank.
• 45:30
  And so he just signaled R2 and R2 shoots the lightsaber.
• 45:33
  He does a flip and comes down. I mean, that that scene hands
• 45:37
  down got to be one of the best in cinematic history.
• 45:39
  Yeah, That's, that's true. That's.
• 45:41
  Fantastic, yeah. Amazing.
• 45:45
  OK. Couple more questions for you.
• 45:46
  We'll let you get out of here. If you went back to school now
• 45:50
  as a freshman or maybe speaking to people that are a college
• 45:53
  freshman right now, does it still make sense to study
• 45:56
  robotics or what would you study differently?
• 46:02
  Yeah, that's that's. Something that we are often
• 46:05
  discussing with with my partner also family, right like what
• 46:09
  would you do when when we were like entering college right now
• 46:14
  or studies right now. I think it's, I think robotics
• 46:17
  is probably still an oak like a a decent bet, right, because a
• 46:21
  lot of this is about design, system integration.
• 46:23
  It's a it's an extremely wide field where you I think you will
• 46:27
  be able to find interesting challenges for humans in the age
• 46:31
  of AI, right. While I think there are like
• 46:35
  many, many, many areas not robotics, right, where I feel
• 46:39
  the world will completely change with AI getting more and more
• 46:42
  advanced and more ubiquitous. Yeah.
• 46:46
  What's a book that? You would buy for somebody or
• 46:50
  give away as a gift. So yeah.
• 46:55
  Good question. So where is where is something
• 46:59
  that I've read, I think maybe two years ago in vacation,
• 47:02
  during vacation, which really like hit me hard, which was like
• 47:06
  this Tony Fadell book about like product management, right?
• 47:10
  It's built. Is that correct?
• 47:13
  So basically that was a book when he was like starting as I
• 47:17
  think at Philips, right? Designing some personal
• 47:19
  assistant, which wasn't very successful, and then moved to
• 47:21
  Apple and did the iPod thing, which was also extremely cutting
• 47:25
  edge at the time, and then moved to iPhone.
• 47:28
  And like, like that's really, you know, I'm talking it's a mix
• 47:34
  of my private life and professional life, right?
• 47:37
  But this kind of building and constructing things and making
• 47:39
  them work, it's really exciting for me.
• 47:42
  At the same time, I really love engineering and advanced
• 47:44
  technology. And this somehow resonated with
• 47:46
  me super hard because it was like both about building a
• 47:50
  product, but it was also about building a cutting edge product.
• 47:54
  So I I can really, really recommend.
• 47:55
  It was very inspiring for me. It's called built by.
• 48:01
  Yes. So.
• 48:02
  It's yeah, I actually have it here by Tony Fadell.
• 48:07
  So that was, yeah, that was an engineer, AKA product manager
• 48:11
  behind the original iPod and iPhone.
• 48:14
  Incredible. OK, awesome.
• 48:17
  We've been talking a lot about movies.
• 48:18
  Do you have a movie you've seen recently or TV show you're
• 48:21
  really excited about? I must say.
• 48:26
  It's a bit difficult recently with two small kids watching
• 48:30
  movies. See a lot of kids movies.
• 48:34
  Yes, I mean, we did watch. Quite a bit, but you can see
• 48:39
  like here that's we watch stranger things.
• 48:42
  So that's actually Lego set upside down.
• 48:45
  I'm I'm I'm really liking 80s and 90s vibe.
• 48:48
  So I also like personally, I I do a lot of do it yourself
• 48:53
  electronics, but also repair old, old or did repair before
• 48:57
  kids all the equipment and cars from the 90s.
• 48:59
  So somehow this vibe of late 80s, early 90s, it's something
• 49:04
  that's that's makes me excited. I would say like privacy, right?
• 49:09
  Cool. How old are your kids?
• 49:10
  Also old one and three, you're watching.
• 49:15
  Stranger Things with a one and three-year old.
• 49:17
  No, no, no, no, no. No, no.
• 49:22
  We tried to watch sometimes shows after they go to sleep,
• 49:26
  but yeah, not very successfully I would say recently.
• 49:30
  Amazing. Now I know Polish charities are
• 49:34
  near and dear to your heart, particularly around medical
• 49:36
  support and help. Talk to us a little bit about
• 49:38
  what's going on in Poland and and the the charitable
• 49:41
  organizations that you support down there.
• 49:44
  So. Like one thing that or how
• 49:49
  Poland is doing. So generally Poland is
• 49:51
  developing very, very well, right?
• 49:52
  And in a way public services and Healthcare is getting much
• 49:56
  better. But still I think there is a lot
• 49:58
  to do right and to bring like in a way Western level of
• 50:02
  healthcare, especially for rare diseases or really serious
• 50:05
  diseases, there is still something to be done.
• 50:07
  And I'm personally like supporting initiatives, right
• 50:11
  which go into financing either equipment or medication for
• 50:15
  those in need where the the the public health for healthcare
• 50:19
  system won't cover that. Amazing.
• 50:23
  It's so cool. Poland.
• 50:25
  How'd they do in the Winter Olympics?
• 50:28
  So they did quite OK. I think especially like ski
• 50:31
  jumping, quite some successes. So that's a, you know,
• 50:36
  traditional thing from from my childhood time that there was
• 50:39
  this guy Adam Maus, who was like basically winning Olympics in
• 50:43
  many tournaments when when I was like whatever, 10 or 12 maybe.
• 50:47
  So it's kind of nice to see a big comeback of, of this crazy
• 50:50
  sport that ski jumping is right? It's basically so crazy, people.
• 50:54
  Flying for like 100? 50 meters in the air with skis
• 50:59
  like a bird and why? Unbelievable.
• 51:01
  Yeah. How do you start that sport?
• 51:04
  You know, it's like, yeah, yeah. I I, I asked myself and actually
• 51:09
  researched this and 1st you start with like really jumping
• 51:12
  like, you know, 5m from a small Cliff or a small Cliff, like
• 51:15
  small ski jump and then you kind of take it from there.
• 51:19
  Yeah, exactly. March in this has been
• 51:22
  absolutely. Wonderful.
• 51:24
  The website is 7 cents dot AI everybody definitely check it
• 51:27
  out. You can see it in the show notes
• 51:28
  below. Wherever you're listening or
• 51:29
  watching this podcast March in what's the best place for people
• 51:33
  to connect with you personally online?
• 51:36
  Yeah, good question. So so.
• 51:38
  I do maintain my small website with a, with a description of
• 51:41
  projects. And also I think LinkedIn is
• 51:44
  always a good chance, right? So, so if you're interested, you
• 51:47
  would like to learn more or maybe you have questions about,
• 51:50
  you know, building AI plus hardware start-ups or industrial
• 51:53
  robotics, right? Just just just drop me a
• 51:56
  message. And I'm always very, very
• 51:58
  supportive and I'm very excited about being able to share this
• 52:02
  experience to others and make them successful.
• 52:05
  Incredible. We will link your.
• 52:07
  LinkedIn in the show notes and then what is your personal
• 52:10
  website? What's the address?
• 52:13
  That's a that's a good question. I hope the domain still works.
• 52:15
  Let me let me check. Yes, it does.
• 52:22
  So it's. My my family name, which is
• 52:24
  pretty complicated. So it's demcheck.com.
• 52:26
  So DYMCZY k.com, we'll put that in the show notes as well.
• 52:31
  I. Haven't I haven't looked into.
• 52:32
  It in a while so I'm I was like OK, I I really hope I paid 2026
• 52:37
  P and not like the card expired or something dealt away.
• 52:42
  Amazing. Thank.
• 52:42
  You so much for the conversation.
• 52:45
  Yeah, thanks. So.
• 52:46
  Much it was a. Pleasure.
• 52:47
  Super interesting and yeah, always cool to talk about the
• 52:50
  future of robotics. This episode is.
• 52:55
  Brought to you by Moberst, the AI first growth engine trusted
• 52:58
  by iconic startups like Calm, Robin Hood, Reddit, Uber and
• 53:03
  Google. Moberst is a digital full
• 53:05
  service, mobile first marketing agency that helps companies to
• 53:09
  scale and become category leaders.
• 53:11
  Whether you're finding product market fit or ready to
• 53:14
  accelerate growth, Moberst delivers creative, data-driven
• 53:18
  solutions that help with scale and efficiency and impact.
• 53:22
  At MO Burst, we're not just guessing what works, we're
• 53:25
  predicting it. As an AI first agency, MO Burst
• 53:28
  uses cutting edge tools to analyze user behavior, adapt
• 53:32
  creative in real time, and uncover hidden growth
• 53:35
  opportunities. The result?
• 53:37
  Marketing that's faster, smarter, and built to convert
• 53:40
  from early stage startups to global brands.
• 53:42
  We've helped hundreds of companies overcome complex
• 53:46
  challenges, unlock new audiences, and optimize every
• 53:49
  metric that matters. At Moeburst, we believe that
• 53:52
  creative thinking, advanced technology, and data are the
• 53:55
  core drivers of success at scale.
• 53:58
  If you're serious about scaling your business with precision and
• 54:01
  purpose, it's time to talk to Moeburst.
• 54:03
  Visit moeburst.com and let's build something category
• 54:07
  defining together.