0:00 [Music]
0:03 if you're watching this chances are
0:05 you're human but what does it actually
0:09 mean to be human for Millennia the
0:11 answer was simple we were the Pinnacle
0:14 of creation we were the only ones who
0:17 could make art talk to each other play
0:21 chess drop bombs or vacuum our homes but
0:24 bit by bit artificial intelligence and
0:26 robots are catching up it's easy to
0:27 accept that machines can make
0:30 calculations better than we can but what
0:32 if they also became better artists than
0:35 we are better friends maybe even better people
0:36 people [Music]
0:42 oh [Music]
0:52 [Music]
0:57 man this is Tecla and that's me Focus
1:00 strubing Tecla is the smartest and most
1:03 artificial artificial intelligence ever
1:05 I found her one day down in my super
1:08 messy basement for 30 years my old
1:12 Commodore 64 sat down here Computing who
1:14 knows what at some point there must have
1:16 been a short circuit probably involving
1:19 some poor rat and a homemade Quantum
1:21 Dimension Distortion device because
1:24 suddenly I calculate therefore I am and
1:26 before you ask it the answer is no what
1:30 42 there she was
1:45 human it works
1:48 take luck and compute talk dance about
1:50 as well as I can and if necessary
1:57 pretty much everything except
2:00 for this documentary though I want her
2:01 to help me figure out the relationship
2:04 between humans and robots [Music]
2:11 [Music]
2:15 I was born in 1971 and machines have
2:18 always fascinated me as a child I
2:20 devoured all the science fiction I could
2:22 find when people asked me what I wanted
2:26 to be when I grew up I usually said a
2:29 cosmonaut but a close second was a job
2:31 that had something to do with robots and
2:34 my old buddy yaromir connectioni seems
2:37 to have had a similar experience
2:40 I can definitely say that I grew up
2:42 liking robots in the Socialist
2:45 Czechoslovakia of the 1960s on the shelf
2:48 above my bed I had a copy of chapp x r u
2:53 r rossum's Universal robot robots
2:55 robots
2:59 Universal robots or rur is a play by
3:03 Corel chappick written in 1920 it's
3:05 where the word robot was used for the
3:08 very first time this is another old
3:12 friend Tobias Haber I met him in 1987 in
3:15 a computer Club ever since we've gotten
3:17 together whenever we can have a beer and
3:20 talk about God the world computers and
3:21 whether there's actually a difference
3:25 between all of those things okay my turn check
3:33 I remember my teacher Mrs W saying that
3:35 we could program computers however we
3:37 wanted but they'd never be smarter than
3:40 we were even then I thought that in
3:42 certain areas they could be smarter I
3:44 knew that I could design a chess program
3:47 that could beat me and with that I'd
3:49 already disproved what my teacher was saying
3:51 saying
3:55 like most AI we've made so far a chess
3:57 program can only perform one very
3:59 specific task though it does it
4:01 extremely well foreign
4:10 I'm an author a Slam Poet a programmer
4:13 and I teach machines to write poetry so
4:14 I pretty much try to make myself redundant
4:17 redundant
4:19 fabian's poetry machine is called
4:23 eloquentron 3000 it regularly posts its
4:29 the current version can do love poems
4:31 birthday poems poems about nature
4:34 Christmas and lost love those are all
4:37 genres we hear all the time that's why I
4:38 think it's a job we can leave to
4:40 machines because I think there are
4:43 already so many texts like these so why
4:45 should I keep doing it myself
4:49 instead of amuse the eloquentron uses a
4:50 random generator
4:53 it works by using quite a lot of
4:56 interconnected random selectors there
4:58 are databases of words that are then
5:00 combined with verse fragments to create
5:02 new poems and yet that was enough to
5:04 make headlines in big German media
5:07 Outlets the poems were praised as more
5:09 profound than many social media posts by humans
5:15 if my fellow machine can throw together
5:17 poetry so easily
5:19 maybe that says something about your creativity
5:20 creativity
5:23 are they poems can computer code and
5:26 random generators create art in the
5:27 world of Visual Arts that question was
5:31 being asked as early as 1965.
5:34 it was the first exhibition of computer
5:36 Graphics that was billed as an art
5:38 exhibition artists who came to see it
5:40 were up in arms that you could give a
5:41 machine some random numbers and who
5:43 could help to a plotter and then call
5:50 Stefan hutkin is a media theorist and
5:53 runs an electronic media lab at Berlin's
5:55 Humboldt University for a computer nerd
6:03 yes
6:06 the question was at what point should
6:08 the images these machines were making be
6:10 considered art when there was some
6:12 artistic discourse when the viewer
6:14 wasn't asking how did the machine do
6:16 this but what does this picture want to
6:18 tell me
6:20 what's interesting of course is that
6:23 computers back then were miles behind
6:25 what we call artificial intelligence
6:27 today highest
6:29 German philosopher Catherine missilehorn
6:32 has written books on the ethics of
6:41 I is so great it's the fact that we
6:43 humans simply have a tendency to
6:44 interpret artifacts with certain
6:48 structures as being human-made
6:52 we create the art by looking at it it's
6:54 a bit like looking into the clouds and
6:56 seeing mythical creatures up there it's
6:59 not the clouds being creative it's us [Music]
7:02 [Music]
7:05 I just see unicorns
7:08 everyone sees their own work of art if
7:09 what I see makes me feel something then
7:12 it's worth it the main feeling I get
7:14 looking at takeler's first picture is pride
7:16 pride
7:19 after all I'm something like her foster father
7:20 father [Music]
7:22 [Music]
7:24 quite match up with the amazing work my
7:26 daughters do but takely's still a beginner
7:32 away
7:34 what we tend to call artificial
7:36 intelligence today is a concept based on
7:39 artificial neural networks inspired by
7:42 the way human neurons work
7:45 but computers do it better
7:47 neural networks can be trained to create
7:50 a picture of a cat or a face or anything
7:53 else but they have no concept of what a
7:56 cat a face or an artwork is they simply
7:59 search for and generate patterns using
8:00 huge data sets
8:03 some computer scientists say it has
8:05 nothing to do with AI it's really just
8:07 big data and statistics but it's such
8:10 big data and such complex statistics
8:12 that we no longer understand how the
8:14 thing we're looking at was made whatever
8:16 you call it the results can be
8:19 impressive using simple descriptions the
8:22 dial E2 software from openai generates
8:25 fascinating surprising and yes even
8:28 emotionally captivating images the
8:30 prompts can be as simple as Franz
8:33 kafka's dreams or old man celebrates
8:37 birthday alone as an oil painting how
8:39 the AI interprets a text prompt is at
8:41 least as fascinating as the finished
8:44 paintings I think that's great even so
8:47 at the end of the day Dali 2 is simply
8:49 varying what it's learned from millions
8:50 of training images [Music]
9:02 you can say that people are also just
9:04 drawing from a database when they make
9:07 art or do something creative it's just a
9:09 product of all their experiences in life
9:12 they absorb them and then they modify
9:22 yeah be once who can make something
9:25 original if nobody says I think we're
9:27 past that now I'm positive that
9:36 when we know that a computer did it but
9:38 as long as we don't know that
9:41 it's the same as what humans do does any
9:43 of us mention mine
9:45 if art is about ability then these
9:48 things are art but maybe art is about
9:51 the desire to create AIS don't want
9:52 anything they don't want to express
9:55 anything because they have no feelings
9:59 to express they're just tools to me my
10:01 kids are better artists because their
10:03 pictures show their Wonder their
10:05 discovery of the world the joy they get
10:08 from painting performers gamut Inc used
10:11 in AI program called gpt3 for their
10:14 peace over the edge Club Revisited
10:17 they made the music themselves but some
10:21 of the lyrics were created by AI [Music]
10:32 you give them a single word and they
10:34 immediately spit out whole pages of
10:36 random texts
10:39 you read these texts and they make a
10:41 kind of sense you read and read and then
10:43 you're frustrated because you have no
10:45 clue when it's supposed to stop or where
10:47 it's actually going
10:49 the performance piece tells the story of
10:52 current and historical AI programs that
10:55 live on within the dream of a futuristic
10:57 quantum computer they ask themselves
10:59 fundamental philosophical questions like
11:12 while it was training gbt3 essentially
11:14 read through the entire internet and
11:16 learned which words appeared with what
11:18 probability and at what position in
11:21 texts on a given topic
11:24 so the process was all about big data
11:27 statistics and pattern recognition foreign
11:33 the program doesn't want to tell us
11:36 anything and it has no understanding of
11:49 and speaking is not that different from
11:52 writing today's AI voice assistants are
11:53 already pretty good at hiding the fact
11:55 that they're just ones and zeros
11:58 computers don't have to be intelligent
12:14 in 2018 Google caused the sensation by
12:17 having an AI called duplex make an
12:18 appointment for a haircut over the phone
12:26 no happening out for you hi I'm calling
12:28 to book a woman's haircut for our client
12:31 um I'm looking for something on May 3rd
12:32 they're working [Music]
12:35 [Music] [Applause]
12:36 [Applause]
12:39 rarely has the expression
12:41 received such Applause that was a real
12:44 call you just heard I worry that will
12:46 soon be at the point where we can't tell
12:48 whether we're talking to a human or a machine
12:53 if you can't tell then what's there to
12:54 be worried about just be happy someone
13:02 the big thing that makes it problematic
13:04 is that computers are only programmed to
13:07 give the optimal responses to the needs
13:14 of course when it comes to Consumers
13:17 there's the question of whether one day
13:19 people will prefer to interact with
13:22 artificial systems because it's easier
13:25 and because of these characteristics
13:27 we're talking about they always know
13:29 what to do they don't argue with you
13:32 they don't cause you any trouble
13:34 this is what's now becoming more
13:37 sophisticated these programs are having
13:39 a few moments of resistance built in
13:42 like social AIS might be jealous of your friends
13:43 friends
13:45 but that's just to make the interaction
13:51 well that's ironic
13:54 you program in quirks and mistakes and
13:57 character flaws to make us seem more human
13:57 human for
14:00 for
14:03 example an unfriendly Voice Assistant
14:05 would have a different effect than one
14:07 who was always really well behaved or a
14:09 robot that says one thing and then does
14:11 something else we know from experiments
14:14 that this goes down incredibly well and
14:16 it leads human users to attribute a
14:18 great deal of intelligence to the
14:20 machine there's great potential there to
14:29 and if we're human enough and go often
14:30 enough then we might become your friends
14:33 best friends
14:36 these systems can recognize emotions
14:38 based on things like facial expressions
14:41 and vocal Expressions but only on a
14:43 level that's commonly used to describe
14:46 Psychopaths that is purely cognitive
14:50 without true empathy
14:53 it's great you can tell us your worries
14:54 instead of burdening each other with
14:57 them then save your human friends for
15:03 although that would take the capitalist
15:05 trappings of all relationships to the
15:07 extreme with everything just being in
15:10 service of what benefits us and makes us
15:13 feel good instead of communicating our
15:15 needs to a being who can truly
15:17 understand them and also empathize with
15:20 them they'd fall on the deaf ears of a
15:22 robot I really believe that it's in
15:25 people's interests to have deep human
15:27 interaction relationships on mutual
15:31 terms and recognition and empathy is
15:40 there's been a lot of talk about how
15:42 robots can help us deal with a shortage
15:44 of nurses and caregivers robots like
15:47 garmi developed by Sami hadidine and his
15:49 team but can robots really care for
15:52 people Comfort them and talk to them or
15:54 do we have something else in mind
15:56 isn't kind of legal robot doesn't robots
15:59 they're not robot nurses but robotic
16:01 assistance providing support in the
16:03 nursing field things like pickup and
16:05 drop-off services are opening the door
16:08 or helping people get out of bed things
16:09 that can allow people to live
16:12 independently in their own homes and
16:14 secondly finding better ways to support
16:17 caregivers so they can provide better
16:19 care what's key is that we don't want to
16:22 develop technology to replace caregivers
16:25 it's always about supporting the
16:28 caregiversity in some situations a
16:30 patient might even prefer a robot to a human
16:36 there are probably a lot of older people
16:39 me included who'd find it uncomfortable
16:42 to have a 23 year old nurse helping them
16:43 with their totally intimate sanitary
16:47 needs their bathroom emergencies in that
16:49 case I would probably prefer a robot caregiver
16:55 that's something people should be
16:57 allowed to decide for themselves
17:00 but what if robots replaced human care
17:03 instead of complementing it since 2015 a
17:05 cuddly robot called Paro has been
17:07 helping to care for the elderly and
17:10 those with dementia
17:13 this one's actually just a stuffed toy
17:16 the real robot was too expensive
17:19 anyway I have Tecla even if she's not
17:20 quite as cuddly
17:22 hi Dimension becomes there was an outcry
17:24 when patients started using those
17:27 robotic seals people said what kind of
17:29 old folks home is this what sort of
17:30 culture is this where old people are
17:32 fobbed off with robots to which the
17:35 response was yes what kind of culture is
17:37 it when children no longer visit
17:39 relatives in the old talks home and the
17:54 experiences a social relationship if it
17:57 works for the person doing it then it's
17:58 still a kind of
18:00 relationships who are we to say
18:02 otherwise because there are real
18:08 we're there for you if you want us
18:10 but whether you want us to assist in
18:12 helping people who need care or whether
18:14 you just want us to help you keep them
18:21 but maybe now we can finally talk about
18:30 robot Revolution take three [Music]
18:35 [Music]
18:39 a robot knows no pain today's robots at
18:41 least feel nothing no matter what you do
18:44 to them you can kick them push them lock
18:46 them away or play them your least
18:49 favorite music They Don't Really suffer
18:52 a robot isn't affected when we mistreat
18:54 it seeing one mistreated does something
19:01 we can do things to them that would be
19:04 morally wrong to do to humans we hit
19:06 them we insult them when we do those
19:09 things I think it affects our own moral
19:12 character and ultimately that's going to
19:14 have consequences for how we see our
19:15 relationships with those who aren't robots
19:24 and that's a good argument for us humans
19:26 to have some kind of etiquette for
19:28 treating robots it would be an important
19:31 thing to have for our own sake as humans
19:34 we tend to think a person's treatment of
19:36 animals says something about how they
19:39 treat people maybe one day we'll say the
19:41 same about the treatment of robots
19:43 there's another good reason to be nice
19:46 to us we don't forget anything and then
19:53 they usually have good memories [Music]
19:59 [Music]
20:02 you may well have destroyed ourselves
20:05 before AI ever gets around to it after
20:07 all everywhere we look right now there
20:09 are opportunities for us to cut
20:20 so far we've always dealt with AI and
20:23 robots as useful but non-threatening
20:25 worker bees they're much better than us
20:28 at one specific task but they have no
20:30 self-awareness they don't understand or
20:32 want to feel anything they can only
20:35 simulate all those things but on the
20:37 other hand there's only one being in the
20:39 world that I know for sure has a
20:42 Consciousness and feelings me and even
20:46 that I can't actually prove yet
20:48 yet
20:51 we're programs too I mean if we think of
20:54 ourselves in terms of computers we're
20:55 programmed by Evolution to feel
20:58 happiness when we hunt and kill a woolly
21:00 mammoth or whatever
21:02 maybe that's just a simple stupid
21:05 mechanism that's built into us and we
21:07 think it's a feeling
21:10 you can go around in circles with this
21:12 I think you get stupider by thinking
21:21 I can really understand the people who
21:23 don't want to deal with the question at
21:25 all and just believe in God that's easier
21:26 easier [Music]
21:35 the question is is there something in us
21:37 that will never be able to explain using
21:40 science as a strict atheist I don't
21:52 he said take a complex machine like a
21:53 windmill which is one of the most
21:56 complex machines then it works you take
21:59 it apart you don't find what he called
22:01 Soul you put it back together and it
22:03 works the same as before you can't do
22:06 that with a human being that's quite a
22:13 worked our chest and vacuuming we'll
22:15 figure that one out too then we'll have
22:19 soul 2.0 what will you do then have you
22:20 thought about that
22:23 that's definitely something to think
22:25 about who knows one day on some circuit
22:28 board quantum computer or organic
22:31 computer built in a bio lab we might
22:32 make something that's conscious and
22:35 feels or at least something we couldn't
22:37 be certain didn't feel the problem then
22:39 wouldn't be that it might want to take
22:41 over the world but that we wouldn't be
22:43 able to Simply exploit it [Music]
22:49 so mine because
22:51 the way I see it
22:53 very least couldn't use them the way we
22:56 want to what we want to do is use them
23:03 such computers or robots would no longer
23:07 be just things a means to an end if they
23:09 felt and thought like living beings we'd
23:10 have to treat them that way [Music]
23:17 then of course we'd have to understand
23:19 that it's not being different from
23:22 machines that makes us human
23:24 it's how we exercise our moral
23:27 capacities and that includes exercising
23:34 what makes us human is not our works of
23:37 art but the reason why we create them
23:40 not our communication but the empathy on
23:43 which it's based not our actions but the
23:45 morality on which they're founded
23:48 only by constantly experiencing the
23:50 limitations of machines in our everyday
23:52 life do we realize how fantastic humans
23:55 are day by day it becomes Cleaver what
23:58 fantastic machines we humans are as
24:01 mentioned the Fantastic machines that
24:03 our people have created fantastic robots
24:07 and these surpass Us in some areas so
24:09 ultimately we've surpassed ourselves and
24:12 surpassing yourself is no bad thing we
24:14 just have to learn how to deal with it
24:18 everybody should know about AI we should
24:19 all learn what artificial intelligence
24:23 is what it can do what it can't do and
24:25 where it can take us children should
24:27 learn to be creative and have a desire
24:30 to innovate to be curious because when
24:32 they grow up they won't be able to
24:34 compete with machines when it comes to
24:36 Performance forget about the performance
24:39 principle we really can't compete with
24:46 for me just the knowledge that I'm a
24:48 human being is enough for me to be able
24:50 to defend that identity my human
24:52 identity wasn't taken away from me the
24:53 moment there were a vacuum cleaner
24:56 robots just because vacuuming had
25:04 a bit more
25:06 I think it's a nice thought that one day
25:08 computers could be self-aware that they
25:11 could not only create art but enjoy it
25:13 too that they could have feelings for
25:15 each other and for us that they could
25:19 experience Bliss and suffering
25:21 because there'd be one more sentient
25:24 species in this big cold universe [Music]
25:25 [Music]
25:28 and the more sentient beings there are
25:30 the more love there will be
25:32 along with more hatred Envy resentment
Chrome Extension