0:00

Brought to you by Toyota. Let's

0:02

go places. Welcome

0:07

to Forward Thinking. Hey

0:12

there, welcome to Forward Thinking,

0:15

the podcast that looks at the future and

0:17

says Yoshimi, they don't believe me, but

0:19

you won't let those robots defeat me. I'm

0:21

Jonvin Strickler and I'm Lauren foc Obam,

0:23

and I was hoping that you would use Yoshimi

0:25

battles the Pink Robots, uh quote

0:28

for this particular episode from The Flaming Lips.

0:30

I'm so pleased you recognized where

0:32

that came from. I think this is like the third

0:35

time that's happened in the entirety of the

0:37

what like two or three years we've been recording. There

0:39

was I think that when I used maps once you got

0:41

that one. To

0:43

be fair, my musical references

0:46

are different than Lauren and Joe's

0:48

musical references. That is generally

0:51

true. Speaking of Joe,

0:53

he is not with us today, as you probably guessed

0:55

from the fact that he did not chime in. Joe McCormick,

0:58

our third co host, will be very

1:00

soon. In the meanwhile, if you

1:02

couldn't tell, we're talking more

1:04

about robots. Yeah, we mentioned in our last

1:06

episode that we were having a

1:08

duo robot conversation this

1:11

week, and in fact, this topic that

1:13

we're going to talk about today, robot security

1:15

guards, was a suggestion from Joe

1:17

himself, and so we hope to

1:19

do justice to the topic, Joe, and

1:22

we look forward to you being back in the studio very

1:24

soon. So let's talk about

1:27

robots and security guards and

1:29

and and robot security guards. Yes,

1:32

because well, okay, we've talked

1:34

many times on the show before about the

1:36

amazing and sometimes strange

1:38

ways in which robots are taking over

1:40

traditionally human work, the

1:42

type of human work that frankly just

1:44

stinks, uh, you know it worked that's

1:47

physically hard or repetitive, or dangerous

1:49

or just boring. Yeah, and all of those

1:52

things are detrimental

1:54

to a person's sense of self

1:56

worth. They can affect job satisfaction.

1:59

Like if your job ends up devolving

2:02

into something that is repetitively dangerous,

2:04

obviously, that can have a profound

2:06

effect on you stress, either physical

2:08

or emotional or both. And even if

2:10

it's just dull, if it's just boring

2:13

and repetitive, then that can have

2:15

an effect on you as well, a big negative

2:17

impact. Absolutely. And okay,

2:19

so so security why is security

2:23

this type of a field, this type of field

2:25

in which we would potentially want to get robots

2:27

into. Well. Well, uh,

2:30

some some background factoids on security.

2:32

According to the U. S. Bureau of Labor Statistics,

2:34

there were about a million human

2:37

security guards employed around the

2:39

United States as of Um.

2:41

They generally make about twelve or thirteen bucks

2:43

an now or on average patrolling and monitoring

2:46

premises for suspicious behavior and

2:48

crime. And they're important because

2:50

even even in this world of ours that is increasingly

2:53

rigged with security cameras everywhere,

2:55

security guards, you know, having

2:57

having a physical person with a physical presence

3:00

can absolutely be a deterrent to crime, right,

3:02

I mean a camera might

3:05

deter a criminal, it might not. And if

3:07

the criminal has decided to go ahead and commit

3:09

said crime, perhaps they've

3:12

got a mask on or something, they feel fairly confident

3:14

that the camera's not going to be able to

3:16

figure out their identity. What's the camera

3:18

gonna do? Not not a whole lot,

3:20

just going to record like it

3:23

usually does the same thing it does every night, pinky

3:25

yeah. Um. And furthermore, as we know

3:27

from every heist movie ever, cameras are

3:29

really easy to trick, uh,

3:31

and very entertaining in the

3:34

ways that you can come up with to do so.

3:36

So okay, um,

3:39

this job is important, but it is

3:41

not an ideal job. It is it is dangerous

3:43

by nature to put your soft and

3:45

fleshy body in the way of a person who's

3:48

desperate or determined enough to steal

3:50

or destroy property and uh

3:53

and furthermore, it can just be lonely and boring

3:55

and patrolling an area over and over again

3:58

isn't really fun. Yeah, so it's already

4:00

fitting two of the three DS

4:02

that are often cited as work that is

4:04

ideal for robots and not for

4:06

humans. That would be the dangerous and

4:09

dull. But depending

4:11

upon where you're doing your security gig, you might even

4:13

hit the third D, which is dirty.

4:15

Dirty meaning not just like, oh, that

4:17

place is poorly maintained. But it

4:20

may be that we're talking about security guards for

4:22

areas like where there might be hazardous

4:24

waste and you don't want people wandering

4:26

into that place. But then you

4:28

put a human being in danger by

4:30

placing that person in close proximity with

4:33

those hazardous materials. So

4:35

what's the solution. Why not

4:37

put some robots there? Robots?

4:40

So here's the thing, Lauren

4:43

went out and started looking at this

4:46

crazy, crazy company

4:48

that has a security robot that's that's

4:50

been active for a couple of years now, particularly

4:52

over in the Silicon Valley area.

4:55

But recently we saw

4:58

some news coming out because he

5:00

has started to use one or maybe

5:02

two of these robots at one

5:04

of their lots. Right. Yeah, we were inspired

5:07

by this news story, and so we wanted to

5:09

to look in to this startup called

5:12

night Scope, which is awesome

5:14

name. Yeah. Night is in like the stabby like

5:17

armored human kind of not not night the

5:19

Holy Grail style. Yeah yeah, not the nighttime

5:22

but the dark dark thing. Yeah. Okay. So

5:25

so in this this startup

5:28

in Mountain View, California, home of Google,

5:30

home of Google and etcetera, UM

5:32

started designing this robot that they called

5:34

the K five and what

5:36

they eventually brought to market is a five

5:38

ft three inch three hundred pound

5:41

shiny white doom egg. Yeah

5:43

that's a good way of describing it. It is conical,

5:47

not comical, although I

5:49

guess depending on what happens to it, that also could

5:52

be the case. But but conical

5:54

in shape and uh three

5:56

hundred pounds, I mean this is this is not a lightweight.

5:59

Yeah, I mean, I am five foot three,

6:01

y'all. For for for reference and and the

6:03

idea. Like, I'm thinking that I could

6:05

easily like ride one of these, like a like a doo

6:07

meg pony, but but I

6:10

probably shouldn't. It

6:13

might start making noise, but we'll talk about that in the

6:15

second Yeah, yeah, okay. So,

6:17

so these things are outfitted with just a whole

6:19

lot of sensors. You've got four high

6:21

definition cameras in there that give it three D

6:23

sixty degree vision. You've got a license

6:26

plate recognition camera, four microphones,

6:28

a thermal camera, a laser range

6:31

finder GPS, and

6:33

a weather sensor um or or an

6:35

environment sensor really you should say that

6:38

can monitor the temperature, the barometric

6:40

pressure, and carbon dioxide levels in

6:42

the area. So what all

6:44

of this does is help it find its way

6:46

around, watch any humans that

6:48

are in the area, detect fires very

6:51

important, uh huh um, watch for

6:53

license plates for potential troublemakers

6:55

like a recently fired employees or

6:57

like a car that seems to be casing the joint something

7:00

like that, and etcetera. Um

7:02

it also has Wi Fi capacity so that it can

7:04

communicate with fellow security bots and

7:07

also with human personnel. Who are either

7:09

remote or on the site, but not you

7:11

know, like poking it right, not right, not

7:13

not walking, not literally walking step

7:15

by step with the robot um

7:18

for furthermore, like like room bas they know

7:20

when they're running out of juice and can go

7:23

head back over to their recharge station. Now

7:25

was interesting to me, beyond

7:27

the technology here, which is already impressive,

7:29

this idea, you know, it seems fairly

7:32

comprehensive for your basic uh

7:34

guarding type of duties. I mean you're

7:36

talking about some surveillance, some recording

7:39

patrolling. Is that nine

7:41

scopes come up with a clever means for generating

7:44

revenue. They're not like, obviously these

7:47

robots have to be mega super

7:49

expensive, but they're not. They're

7:51

not just saying, hey, come look

7:53

at our catalog and you see this

7:56

dollar robot. How many would you like to order?

7:58

Yeah? No, they they have a rental

8:01

business model. So uh

8:03

so the official company line and how much

8:05

they cost to rent is that you should totally

8:07

speak with a sales representative to find out a plan

8:09

that works best for you. But

8:12

but news reports have listed it at about seven

8:15

dollars per robot per hour, or

8:18

about sixty dollars a year for seven

8:21

surveillance. Right, So if

8:23

you think, uh, seven dollars per

8:25

hour for a robot, like that's

8:27

probably less than what you would

8:30

pay a human person, and

8:32

in fact, according to averages, it is.

8:34

So in that case, you've got night Scope

8:37

having a sales pitch saying, well,

8:39

if if your needs don't

8:41

require an actual human being to be

8:44

there all the time, all the time, then

8:46

perhaps this robot would be a way that not

8:48

only could you maintain security at your

8:50

space, but you could save money in the process.

8:53

Yeah, and uh and and under under this kind

8:55

of concept, they are in use by

8:57

a few tech companies around the Silicon

9:00

Valley area, like like qual Common and Uber

9:02

we mentioned, and they

9:04

were certainly in use for a demonstration at Microsoft.

9:07

I'm not sure if they actually Yeah.

9:10

Well, what's interesting to me is when I started doing research

9:12

on this, I keep coming up with articles

9:15

from different time periods, like two

9:17

years apart. They were like and every time

9:19

it was like people had just discovered them

9:22

right right, and I had known I had only

9:24

seen them right now. So and I read

9:26

a lot about robot I'm

9:28

like, how did I miss this? They're glorious, to

9:30

be fair, even people in Silicon

9:33

Valley have missed it because every time

9:35

there's another report about these robots being put

9:37

in place by someone, it's like, this

9:40

is amazing, it's brand new, that

9:42

happened two years ago. There's a hashtag on Instagram.

9:45

I don't anyway, Yeah, that is hashtag

9:47

security robot. By the way, if

9:49

you'd like to look it up and see some pictures and videos

9:52

of it interacting with humans. Um, so

9:54

let's say that you rent a few of

9:56

these. How do you set it up? How

9:58

does it know where to patrol? Well,

10:01

you lead it around with a wireless

10:03

controller to help it build out a basic

10:06

map of its patrol roots. That makes sense because

10:08

it sounds adorable. It wouldn't magically

10:10

have the information stored in its

10:12

data banks. It has to learn clearly, So

10:15

yeah, it's it's kind of like taking an

10:17

employee on a tour, right exactly,

10:19

just your giant metal employee.

10:22

Well, and the other neat thing is because it's a machine,

10:24

it doesn't forget unless its memory gets

10:26

corrupted, so exactly will always remember

10:28

where everything is. Yeah, and so, and

10:31

by observing its surroundings, it starts to

10:33

learn, you know, where it can and cannot roll

10:36

because it is a rolling machine, walking

10:39

machine. So you can defeat the security

10:41

robot by with a short curb or

10:43

yeah, exactly, just a just a slight ledge.

10:45

It's it's kind of got the classic

10:48

dollic uh

10:51

weakness. Although I know, I

10:56

know that today they can fly, but

10:58

back in the day they couldn't know. And

11:01

and the first time that a Dollek said elevate and

11:03

went up some stairs, like, I was terrified.

11:06

I was genuinely upset. Spoiler

11:08

alert for first season of Dr Who

11:10

reboot season nine. I guess of the

11:13

technically, but whatever at

11:15

any rate. Um, yeah, so so you so you lead

11:17

it around to to show it where it can

11:20

cannot go. It will start building up its own

11:22

personal database of its surroundings

11:24

there. And furthermore, you can program in GPS

11:27

coordinates to give it like an outline of

11:29

like like a fence almost. Yeah, so it's like

11:31

it's a very invisible fence that the robots

11:33

not going to go beyond. All right,

11:35

So all right, I've decided

11:37

to rent my security my RoboCop

11:41

and and he says, your move, creep. No wait,

11:43

no, I put him. I put him out there, uh,

11:46

patrolling the area that we have determined.

11:50

Let's say that it detects something

11:52

hinky is going on. What happens next?

11:54

Not a hinky thing anything but a hanky thing. Um,

11:57

it phones home, and by home, I

11:59

mean whatever humans are on security

12:01

team at that current moment um. You can have an

12:03

alert different people during different shifts, and

12:05

anyone who gets an alert can tune into

12:08

a live stream of what the boat is

12:10

seeing and hearing. Okay, So

12:12

in other words, you whether it's you've

12:14

got a team that's on site, but they

12:16

are maybe they're overseeing the entire

12:19

area. Like if you're talking about a really big

12:21

area, you might have multiple robots to like

12:23

on different floors of a building for example. Oh absolutely,

12:25

Or you might even have I imagine

12:28

like a like a larger security call center

12:31

that that's servicing many different robots

12:33

over different businesses. Right, that would make sense,

12:35

and you would have certain teams assigned to specific

12:37

robots. Therefore you would get that notification.

12:40

So all right, let's say that, uh,

12:43

we've got the security robot. Clearly,

12:46

this is something that's meant to interact with human

12:48

beings, whether it's someone

12:51

that needs help or it's someone

12:53

who's doing something they shouldn't be doing so,

12:56

how how does the robot actually perform

12:59

those interactions. Does it just like say like

13:02

please don't do that? Well,

13:05

yeah, yeah, because because on on the ground level, you know, like

13:07

it's it's going to phone home if it senses a

13:09

problem. But um, but it's

13:11

it's the humans in the in the security

13:13

team that are going to actually choose

13:17

what what happens, like basically

13:19

the robot disengages. Essentially,

13:21

the robot is programmed to not interact

13:24

with humans. Um, I mean it's it's which

13:26

is a strange feature because and I think

13:28

that the team put really quite a lot of research

13:30

and development into the physical

13:33

design and the response design

13:36

of this robot um too, to

13:38

to make it this object that can

13:41

interact when it has to. But that really

13:44

isn't meant to It's not it's not

13:46

interfering with someone. It's not it's

13:48

not intimidating someone who is

13:51

perhaps just they're

13:53

under normal circumstances or

13:55

perhaps they need help and they're coming to the robot

13:57

to get help. And because the robot actually

13:59

can can communicate back to yeah,

14:02

there's there's a button for that. If you need help, you

14:04

can push a button. It's essentially an emergency

14:06

call button. Yeah, exactly. So you don't

14:08

want to design your robot to look like, you

14:10

know, like the Queens and aren't Killbot two thousand

14:12

with spinning saw blades

14:14

all over. Yeah. Yeah, You're not going to go up to that kind

14:16

of robot for help. I mean also, you gotta run

14:18

from that robot screaming help. Yes.

14:21

Furthermore, if if your if your kid wanders up

14:23

to it, that's it's bad time. It's probably

14:27

you don't want so. So you know, you

14:29

wanted to make it intimidating enough to

14:31

deter crime, but you don't want to make it unfriendly,

14:34

cute enough that it doesn't scare people, but

14:37

not so cute that people are twisted,

14:40

lee willing in, or motivated to mess

14:42

with it, as we have seen

14:44

before in studies of human robot

14:46

interaction. Uh remember that

14:48

that hitchhiking robot, the one that successfully

14:51

got all the way across Canada and then

14:53

maybe two days into the United States before

14:55

it was stolen and disassembled. Good,

14:59

good job, Pennsylvania. That tells you

15:02

so much about the different cultures

15:04

of Canada and the United States. Or

15:07

yeah, okay. There was also a study in Japan

15:11

of how people would interact

15:13

with a robot that was designed to to help

15:15

the elderly shop Yeah, good

15:17

good robot. Kids messed with it even

15:20

and perhaps because it was saying things like

15:22

somebody helped me or ouch that hurts.

15:25

But I imagine in Japanese there is something

15:29

tragically comical about a robot

15:31

crying out saying somebody helped me, because

15:35

you know, it's not really feeling anything.

15:37

Yeah, it's not in pain, and that

15:39

kind of makes it funny, and that kind of makes

15:41

you want to mess with it more. I mean, I'm

15:44

a softie, so I have a feeling that I would

15:46

totally feel empathy for such a machine,

15:48

but I can also see the

15:51

perverse appeal of messing with

15:54

it further. Right, Yeah, there's there's really

15:57

interesting and probably terrible

15:59

psycho oology at work here.

16:01

Um So, so the team,

16:04

it seems like what was aiming to make the K five

16:06

like clearly inhuman

16:09

like, like clearly a a robot

16:12

and not attempting to to tug

16:14

on your heart strings. Right, it doesn't have big puppy

16:16

dog eyes no no, no, no like

16:18

that. Yeah, but also obviously

16:20

autonomous and obviously intelligent.

16:23

Um So, if it's approached

16:25

by a human person, it will turn

16:28

turn to watch you with all of its

16:30

big front sensors because it kind of kind

16:32

of has a front in the back. Um, if you corner

16:35

it, it will stop moving. Just I'm

16:37

pretty boring anyway, Why would you want to mess

16:39

with me? I pounds? Where are you gonna

16:42

come on? What are you gonna do? Stop? Um?

16:44

If you keep it cornered or if you start poking

16:46

at it, it will chirp at you and

16:48

and also phone home like like a low level alert

16:51

like hey, And then you can actually

16:53

have a person say something like hey,

16:56

knock it off, we can see

16:58

you. Yeah.

17:00

Yeah, and if you if you try to mess

17:02

with it, it'll sound an actual alarm

17:04

like like not like a friendly like hey, more

17:07

like a yeah, more like more

17:09

like imagine the worst fire alarm

17:12

you've heard at the loudest level,

17:14

something akin to that, because

17:16

I have seen uh it

17:18

described as ear piercing

17:21

Yes, uh so so yeah

17:23

and and and then then that will send

17:25

like an official alert to to the

17:27

team managing it, and uh and

17:29

the team managing it can I'm pretty

17:32

sure, like I like I wasn't clear, but but it sounds

17:34

like you think that this is the case that the team can

17:36

like speak to you through the robot. I mean,

17:39

if you have if you have a help button, I

17:41

would imagine you have to have some sort of communication

17:44

back or else. All I would do is have like

17:46

an elevated sense of this button

17:48

is doing nothing and we just beak

17:51

behind the curtain. Laurenen Lauren wrote, and

17:53

I recorded an episode of brain Stuff

17:55

about placebo buttons. That's one case where you

17:57

don't want a placebo button. No, No, that's emerge

18:00

as they call. You want a real effective button. Yeah,

18:02

that would be very, very upsetting. So my

18:04

guess is, while I didn't see anything specifically,

18:08

uh yeah, common

18:11

sense would dictate there has to be some sort of speaker

18:13

through which people could could talk. And

18:16

I would imagine that you would that would

18:18

become useful both in a security feature where

18:20

you say, hi, we're you're under

18:22

surveillance, don't do that, go

18:25

away, or we

18:27

we heard you were sending police

18:29

or ambulance or whatever the emergency

18:31

might be to your location right

18:33

now, and we'll stay on the phone with you that kind

18:35

of stuff. Yeah. Um, but but there's also

18:37

a list of things that will will absolutely

18:40

not do. Yeah, it won't physically interact with a

18:42

person. It's not gonna it's not gonna roll

18:44

up and be like, hey, back off, dude. I

18:46

likely

18:50

did the universal fronting

18:52

like pull me back, coming back. Yeah,

18:55

that's not going to do that. Um. The

18:57

closest you will get to any sort

19:00

of physical interaction is that alarm,

19:02

which is powerful enough for you to perhaps

19:04

and sound is technically a physical

19:07

phenomenon. So I guess by then you

19:09

could say that's how it gets physical with you. But it doesn't

19:11

like like, it doesn't have a little

19:13

zapper that comes out allah r two D

19:15

two. No, it's not weaponized in any way.

19:18

It's it's not it's not more about that one in just

19:20

a moment. It's not going to mess with you back.

19:23

Um. And I'm delighted

19:25

by by the existence of this thing,

19:27

a little bit creeped out. Um. You

19:30

know, the fact that it can look me right in my beady eyes.

19:32

You know, I'm not entirely sure how I feel about that.

19:34

I also the fact that they

19:36

one of the owners of the company or one of the creators

19:39

of the robot, in one interview, said like it's the

19:41

absolute minimum height that it could

19:43

be in order to Um, to look around

19:45

a parking lot, and I was like, absolute

19:47

minimum.

19:50

You can't be any shorter, Lauren, you

19:52

will be lost in every parking lot. Um.

19:56

Well, one of the things that makes me think about is that, Well,

19:59

on the one hand, and like you

20:01

get into the discussion about, yes, this is

20:03

taking a job that is not the

20:05

greatest job for a human being, But on the other hand,

20:07

you're like, well, there are human beings who are depending upon

20:09

jobs like these in order to make a living. So there's

20:12

there's two sides of that point. You

20:14

would hope to be able to come up with a means,

20:17

and it's beyond the scope of

20:19

nights scope. It's beyond their scope to

20:21

come up with the means, But you would hope you could

20:23

come up with some way of getting

20:26

those people who would otherwise be be out

20:28

of a job into a different line of

20:30

work or some other related field

20:33

so that they could continue to make a living and contribute

20:35

to society while also not being

20:37

put in physical harm or in

20:39

a situation that's going to be

20:41

soul sapping because it's so dull and repetitive.

20:45

The robot is not going to care, so

20:47

so it's up to us to care. Yeah, I also

20:49

I like the idea of a robot being able to

20:51

to record everything, uh

20:55

perfectly, like with video and audio.

20:58

Uh. Once upon a time, Laren,

21:00

before I worked at How Stuff Works, by

21:02

about ten years, I had

21:05

attempt job working

21:07

for a security a private security firm.

21:09

My job was to transcribe security

21:12

officer reports, handwritten

21:15

reports, so that they could be filed in the computer

21:17

system. And that

21:20

was not the best or most

21:22

rewarding job I've ever had. The people who were

21:24

working there were perfectly earnest

21:27

and hard working, and I

21:30

give all the credit to them. They were not

21:32

writers, and sometimes deciphering

21:35

what they were writing, not just like physically deciphering

21:38

what a word was, but trying to figure out what

21:40

they've meant. Um,

21:43

I'll have to tell you the story of Cookie after

21:46

this podcast. I tell it now, but it would be

21:48

such a huge tangent that it would be a waste of time,

21:50

so I'll tell you afterwards. But

21:52

but you wouldn't have to worry about that because the robot

21:54

would be recording everything. So that

21:57

of data and having a

21:59

cont tenual stream of that data

22:02

at a certain point becomes a little bit of a big data problem.

22:05

Um, But it's probably a good problem

22:07

that many companies would love to have, well, especially

22:09

if you have to pursue a legal claim

22:11

against someone, because you're not relying upon

22:14

uh eyewitness

22:16

account or a security person's account

22:18

of what happened. For a stationary camera's footage

22:21

right where you might not have a good view, you've got

22:23

a three degree view, a mobile

22:26

view, because they can move to wherever the problem

22:28

is or somewhere where has a better

22:30

line of sight of where a problem might be and

22:33

have a recording of that thing that is

22:35

incredibly valuable, well beyond like

22:38

an eyewitness account, which we know is

22:40

not entirely or even largely

22:43

reliable or scientifically. Yeah. Absolutely,

22:45

um. But but this, this night Scope

22:48

K five is not the only security

22:50

bot out there. No, I found a couple of others.

22:53

There's the Avatar three security

22:55

robot from Robotex.

22:58

It's t e X for the into that

23:00

word. But it's twenty four by

23:02

fifteen point inches by six point

23:05

fourteen inches, which is about sixty two by

23:07

thirty nine by fifteen point six centimeters,

23:10

so a little bit more

23:12

than two ft long, a little bit more than

23:14

a foot tall, and bigger than

23:16

a bread box half Yeah,

23:18

yeah, and uh it looks like

23:20

a like a little tread robot, Like it's

23:22

got treads to propel it, so don't

23:24

we Well, it's got wheels, but wheels to turn the tread

23:27

like so to treads kind of like

23:29

a tank. It's got a little kind of

23:31

glass dome on the top that has all the

23:33

cameras and sensors on it, and a pull

23:35

for an antenna to connect to

23:37

WiFi. This one is

23:40

under manual control. It's

23:42

not intended to be autonomous, but it

23:44

would allow someone to patrol

23:46

an area without having to leave their

23:49

they're nice air conditioned security room,

23:51

for example. Oh yeah, you're to check out a situation

23:53

that could potentially be dangerous. Yeah, if

23:55

it's if it's something where, you know, maybe someone

23:58

spotted a box and you're like, well,

24:00

let's get a closer look before we call in

24:02

any authorities. Yeah, because

24:04

it may just be that, oh, someone accidentally

24:06

set their stuff down while loading things into a car

24:08

and then drop off for getting it. Or it

24:10

could be that it's clearly a package of ramen.

24:13

Yeah, then let's not let's

24:16

call the swat team. Uh. So it's

24:19

a little different, it's not it's not like the

24:21

K five and that's the K Fi've been semi

24:23

autonomous, mostly autonomous, really, But

24:26

then you have the a mott a n

24:28

bot from China. Now, this

24:30

one is a little shorter and

24:33

and and about half the weight of

24:35

the K five. It's one point five meters tall,

24:37

which is about five ft and ways seventy

24:39

eights are about a hundred seventy two pounds.

24:42

It's also cone shaped. In fact, it

24:45

looks a lot like the K five, to

24:47

the point where I start to wonder if this is one of

24:49

those cases where one person

24:51

maybe possibly copied down

24:54

an idea of someone else had. It could very

24:56

well be that these are parallel forms of development

24:58

that have nothing in connection with each other. Happens frequently

25:01

remarkably similar designs. However,

25:03

um to the point where when I looked at the first

25:06

picture, I thought, well, why do they have a picture of the K five

25:08

on here? And then I was like, wait, that's

25:10

not the K five. That's central and but

25:13

um so uh.

25:16

It is meant to patrol areas

25:18

prone to civil unrest, according

25:21

to Newsweek, at any rate, and it is autonomous,

25:23

and that it can patrol an area, has

25:25

obstacle avoidance technology,

25:28

and it is able to go back to like a recharging

25:31

point in charge itself. But it

25:33

is also quote equipped with weapons,

25:36

although to be fair, those weapons must

25:38

be manually operated by someone.

25:40

So it's not like the robot just decides

25:43

to hit you. Yeah, that's that's a much

25:45

better option. Someone controlling the robot has decided

25:47

to hit I guess, depending upon who's controlling

25:49

the robot. The weapon, by the way, is an electrically

25:52

charged right controlled tools, so essentially

25:54

like a stun gun or a cattle prod

25:56

something along those lines. So it is like

25:58

the zapper from R two D two popping out

26:01

and giving someone a jolt of electricity.

26:04

But if that ever does happen to you, if

26:06

you are ever zapped by

26:08

an and bot, it's because

26:10

somebody controlling it doesn't like you

26:12

or doesn't like what you're doing. It's

26:15

not because the robot didn't. Yeah,

26:17

unless the robots malfunctioning in some way. If

26:20

you see an and butt in China and

26:22

there's a little protrusion extending

26:24

from it and sparks are coming off, just walk the other

26:26

way, yeah, don't. You

26:29

don't need to go in that direction anymore.

26:31

There are other places they're China is big there's

26:33

a lot of stuff to see. Yeah, so anyway,

26:36

um, I thought that was interesting. But there are of course

26:39

tons of robots that sort of fall

26:41

into the spectrum of security. But

26:43

but most of them are really not meant

26:45

to be security robots. They are robots

26:48

that have features on them that can

26:51

be part of security system. Typically

26:54

we're talking about cameras, so vacuum

26:57

robots that happen to have a camera on them. While they

26:59

don't necessarily need the camera for doing

27:01

the vacuum job, but it allows you to go

27:04

through an app and see what's going on in your house

27:06

while the vacuum robot is active

27:09

and it acts as sort of a surveillance camera.

27:11

So really it's it's not so much a security

27:13

robot. It's more like just a webcam. Well it's

27:17

it's a robot that does

27:19

Yeah, it's it's on the it's on the light

27:22

end. It's in the spectrum. But

27:25

this is just where we are. Now, where

27:27

are we going? This is forward thinking think

27:30

about the future. Uh yeah,

27:32

well, I mean okay, like

27:35

we were saying earlier, this this could

27:37

potentially if we if we follow

27:39

this out into having full fledged

27:41

security robots, we we were gonna

27:43

have to deal with a with a with a job deficiency

27:46

UM and and Night Scope says

27:48

themselves that that robots like the K five are

27:51

not intended to replace human

27:53

security guards certainly not yet UM,

27:56

but rather to keep them out of harms away as much as possible,

27:58

and to arm them with better ADUTA about

28:00

the environment that they're guarding. But

28:04

but what if we do create robots that

28:06

are sufficient enough to replace replace

28:08

HUM human staff Like they they have the

28:11

features that will allow them to do a security

28:13

guards job without the

28:15

need for the team

28:19

team right right now, nothing to to make the decisions

28:21

or to come out and zap people themselves. It would

28:24

be tough, uh tough

28:26

for many many reasons. One

28:28

of the big ones that we've said it other

28:30

episodes. There's some things that humans can

28:33

do really well, like naturally well we you

28:35

learn how to do it as a kid and you do it for the rest

28:37

of your life, and robots are

28:39

not good at those things. So,

28:41

for example, the K five isn't

28:44

really great at staying upright on uneven ground.

28:46

There was one or one article

28:49

I read where they talked about as they were chatting

28:51

with a representative in the background,

28:53

they saw a K five get too

28:56

close to the edge of like a curb and

28:58

tilt over, and it

29:00

doesn't have arms. It can't push itself

29:02

back out. It can't. Yeah, so then you have to pick up

29:04

a three pound robot and put it back on its

29:06

base. Um, not

29:09

necessarily easy to do. And the Darbo

29:11

robot challenge that we talked about in

29:13

a previous episode, Yeah, the best

29:15

people in the world working

29:17

on robots that can do a multitude

29:20

of tasks are still pretty

29:23

crap at making robots that can walk around.

29:26

Yeah, like that. It's hard to make a robot

29:28

that can climb stairs or open a door and

29:30

step through things that

29:32

that most people find you

29:34

know, at least, if if not effortless,

29:38

at least not so challenging.

29:40

It's impossible, right, but

29:42

there are robots that literally like you can watch that

29:44

that video clip of the robot fails

29:46

of the Darba challenge, and usually

29:49

it's set to yakety sax, which makes hilarious

29:51

and also sad at the same time. But

29:55

it's it's crazy to see the little things

29:57

that robots just find incredibly

29:59

challenge JNG. And you start when

30:01

you start to think about you realize, yeah,

30:04

replicating what humans can

30:06

do creating a machine that can operate

30:08

within a human world, a world

30:10

that we have shaped, that that we able

30:12

bodied more or less average

30:14

sized humans have shaped. Right, It

30:16

is not easy to do that. Um.

30:19

So Also we have to point out security guards would

30:21

likely need at minimum the basic skills

30:23

of a typical, able bodied human, and

30:26

maybe more than what just

30:28

a typical person would be capable of doing, depending

30:31

upon the type of security guard and the type

30:33

of gig right, and and plus in addition

30:35

to that, if you don't have that team of people

30:37

that it can phone home to in order to make decisions

30:40

for it, you need the capacity to autonomously

30:43

decide when, when someone's a threat,

30:45

how much of a threat they are, and

30:47

then how to stop them. Um. Computers

30:50

only just learned how to identify cats.

30:52

Yeah, I'm not sure how long

30:54

it's going to take us to get to a

30:57

design of of that level

30:59

of decision makeing and intelligence until

31:01

I can encounter a robot that not only

31:03

knows what what cats are,

31:06

but can tell you don't go see the show cats,

31:09

it's really not that great, and furthermore,

31:11

don't scritch its belly immediately. Cats.

31:15

Yeah, there are a lot of like

31:17

a lot of subtle things. I

31:19

think exactly that if a robot

31:22

isn't able to tell me what those are, I'm

31:24

not going to trust it to be guarding

31:26

something of supreme importance,

31:29

especially in a way that is at

31:32

least in a way that's not going to hurt somebody,

31:35

Right, which brings us to the next point,

31:37

the idea of security guards

31:40

occasionally have to use some form of

31:42

force in order to protect a person

31:44

or protect property. Yeah. Yeah, we would

31:46

be talking about designing a robot

31:49

to physically incapacitate humans,

31:52

and instinctively that seems bad.

31:54

Yeah, I mean I

31:56

mean at the very least, like potentially hazardous

31:59

and litigious. Yeah, absolutely,

32:02

I mean, like that goes

32:04

back to our discussion from European Union

32:06

and liability. Who's liable?

32:09

Well, if you've got a security robot

32:11

zapping people, I mean, even if they've done

32:13

something wrong, there's there's

32:16

some liability issues there, and it does seem

32:18

to go against those laws of robotics we talked about

32:20

in our last episode do not harm humans.

32:22

Yeah, that's a big one. It's right, there's number one.

32:24

In fact number one that the rest of them

32:26

are kind of all based on number one. So yeah,

32:29

and there exactly

32:31

so. And again these are laws

32:34

that robots are born with. But

32:36

these are the sort of ideas we've had of saying,

32:39

if we're going to go down the road of

32:41

building robots, we probably

32:43

want to make sure they're not going to hurt us. That

32:46

seems like a pretty wise thing to build

32:48

into them. Yeah, and I do think that

32:50

that part of that,

32:52

that intrinsic instinctive

32:54

sense that we have about creating robots

32:57

that could hurt people is based in

32:59

science fiction. Um, is based

33:01

in all of the fun

33:03

exploding movies that we see where

33:06

all Arnold Schwarzenegger has a metal skeleton

33:08

and Any's stomping around and terrifying

33:11

the living daylights out of everyone and all

33:13

of that, all of that kind of stuff that that

33:15

is so far away from being any kind of

33:17

reality. Even if you do create a

33:19

security robot that can zat people, UM,

33:22

that's that that robot is not are Schwartzenegger.

33:25

Uh, it probably doesn't even have an

33:27

Austrian accent. I can't

33:30

I can't be positive. I don't know. I don't know what they're

33:32

up to in China. Kind of want them all to have it now.

33:35

But but yeah, but you know, it's the idea

33:38

of a robot running a muck is

33:40

pretty pretty unlikely. Yeah,

33:43

really, yeah, I mean it's yeah,

33:46

we we have a ways to go before that

33:48

that becomes before Paul Paul McCartney

33:50

butt is a reality. Yeah yeah,

33:55

but okay, so I don't know, you know,

33:58

it's like, if you're really just designing

34:00

something like this to prevent crime,

34:03

which is bad. Crime is bad. And

34:06

if and if it's a robot that's freely

34:08

good at making decisions and it definitely

34:11

super never kills or maims anyone

34:14

accidentally or on purpose, those would

34:16

both be that would also fall into

34:18

the realm of bad. Um.

34:21

If all of that is true, then

34:24

then it would be a good thing to have a

34:26

security robot. Right, So we just need

34:28

a robot that can have like a whole bunch

34:30

of arms and it just very

34:32

gently hugs the criminal and

34:35

holds the criminal there like they're

34:37

all all arms are soft, so as

34:39

the criminal is struggling to get away, he or she

34:41

is not actually hurting him or herself and

34:45

yet can't get away because robot

34:47

also weighs like seven pounds,

34:49

just like just just a giant panda robot.

34:52

Is what I'm imagining that kid

34:54

that can hug a criminal into

34:56

submission. Yes,

35:01

I approve of this plan. I

35:03

will pitch this series to sci Fi

35:07

call us. Yeah, if

35:09

we can get this on, we can get this on sci

35:11

Fi. I think we got a future in

35:13

uh in science fiction programming.

35:16

No, no, I mean if if we can create

35:18

this, this panda security, but you know, it's

35:20

it's it means fewer humans being

35:23

exposed to risky situations, which is always

35:25

kind of one of those like like golden

35:27

standards of how robots will do us

35:29

good in the incredible future. Yeah,

35:31

so there's definitely a place

35:34

for it. The question is to what extent I

35:36

think. I think the K five is a great implementation

35:39

of this because it is taking

35:43

what appears to be a very responsible

35:45

approach to incorporating robots

35:47

into the realm of of security,

35:50

um without causing at

35:52

least initially anyway, at least at least

35:55

in no obvious way risk

35:58

or danger. Yeah. So, uh,

36:01

I can't wait to

36:04

revisit this topic in like ten

36:06

years and talk about remember our

36:09

innocent days when the robots

36:11

were being friendly and when we thought

36:13

that panda robots were a great plan.

36:15

The other day I went and by myself

36:18

a Soil and Green bar and it

36:20

was a little slow reaching for my money,

36:22

and um, well that's why I got the cybernetic

36:24

leg. Now coming

36:27

up next, we're gonna talk about the future

36:29

Soiler Green. This episode

36:31

brought to you by soil It. No, I

36:34

mean that's I'm hoping that's not going to be

36:36

the way this show isn't ten years but we'll find

36:38

out. Um, maybe thought

36:40

that would be a really interesting accent for you to spontaneously

36:43

developed. Well, it may be one of those that's mandated

36:45

by our robotic rewards. Like

36:48

Jonathan, you have to have a weird cartoonish

36:50

accent for the rest

36:52

of your life, at least when you're on microphone.

36:55

Um. Alright, So guys, if you have any comments

36:58

about this episode, or robots

37:00

in general, or even just some other topic

37:02

that you want to know, like what is that going to be like

37:04

in the future, right to us. Let us

37:06

know what you think we want to hear from you.

37:08

You can send us an email. The address is FW

37:12

thinking at our stuff works dot com,

37:14

or you can drop us a line on Twitter or Facebook.

37:17

We are FW thinking on Twitter, or

37:19

you can search f W thinking on Facebook. Our

37:21

profile will pop up. You can leave us a message and we

37:23

will talk to you again. Really. See

37:30

For more on this topic and the future of technology,

37:33

visit forward thinking dot Com,

37:46

brought to you by Toyota. Let's

37:48

Go Places