Player is loading...
Artificial intelligence and quality of life
H. Bourlard, Idiap Research Institute
Saturday, Sept. 11, 2021 · 4:30 p.m. · 17m 03s
Embed
Transcriptions
Note: this content has been automatically generated.
00:00:00
if a hole are you have control
00:00:09
thank you very much
00:00:13
i know um
00:00:17
and uh i uh scientists but i like to give presentations which are rather a controversial
00:00:25
um a a more accustomed to speaking in english than french and uh some of my slides or in english
00:00:32
so i'm going to talk to fifty minutes about artful artificial intelligence
00:00:36
will the quality of a life i hear a lot about it um
00:00:41
uh revolutions in the field of a i on the radio and t.
00:00:44
v. and turned social media back in nineteen ninety president george bush did cleared
00:00:52
the next decade as the decade of the brain and artificial intelligence
00:00:58
tens of billions of dollars of being injected into a high since nineteen ninety everyone
00:01:04
back in nineteen ninety i was a young research myself we were very
00:01:08
proud to display this a patch when we went to conference is this um
00:01:14
a lapel badge we all displayed this uh uh edge
00:01:20
um we were convinced that we were part of the revolution of the human brain
00:01:25
the decade of the human brain nineteen ninety to two thousand
00:01:31
george bush invested billions of dollars in that invent adventure in
00:01:35
january two thousand and ten and my keep published a report
00:01:42
entitle decade of the brain and i invite you to read it
00:01:48
it's an update of a plate and a yeah
00:01:53
of brain and a high research what's that mean that this
00:01:59
doesn't necessarily mean that researchers did that to a bad job it
00:02:02
just means that whenever you are dealing with a a high and the brain what should not underestimate the complexity of the task
00:02:09
uh these uh extremely complex process is and that's what i will try to talk to you about today
00:02:17
no one sentence which i like an which is often repeated is that
00:02:22
there is no intelligence in artificial intelligence country to what we read in the newspapers every day
00:02:28
it's not the single slide to sign of intelligence in artificial intelligence
00:02:35
oh
00:02:37
but um as a for a bunch of science is being a developed recently which
00:02:41
is very very sound very sound i will come back to us like to i will
00:02:45
tell you what we know how to do today what we don't yet know how to do
00:02:49
when we ask ourselves questions about the quality of life when we ask questions about the the subject
00:02:54
some to says well i work on a i and i called into question but i'm worried when i see that a lot
00:03:00
of people are asking questions about a i mess it myself this may well be the first time in the whole of humanity
00:03:08
ah
00:03:12
um promiscuous um uh uh still the
00:03:17
file from his father's use and who
00:03:21
gave prior to humanity and since that
00:03:24
h. senses never being called into question
00:03:29
sounds was understood is paying for the benefits of humanity it's
00:03:33
only recently that people started to ask questions about science itself
00:03:39
it's part from that since time immemorial since promiscuous
00:03:43
just being considered that science was at the service of society and the quality of life to where of course if you
00:03:49
exceptions we all know about galileo who was condemned by
00:03:53
the inquisition and so on but that's slightly different nuclear bomb
00:03:58
we started i'm asking questions about the relationship between science and a society in that respect
00:04:05
oh this is a prime yeah in the history of a human to even madison
00:04:09
is being called into question we had already always taken for granted the fact that madison
00:04:14
was working for the benefit of a humanity but now
00:04:18
people are coming up with conspiracy theories about a medicine why
00:04:24
there's no intelligence in artificial intelligence but just sound science
00:04:30
have to realises no intelligence in a i'd but just sound a science
00:04:37
and which we'll mustering more more on which result from the convergence of several
00:04:40
technologies at the physical level the development
00:04:43
of infrastructure memories a calculation capacity et cetera
00:04:49
all this is linked to convergence at the level of programming algorithms mathematics statistics
00:04:58
there's no miracle here this is scientific progress
00:05:02
and then there's what i would call the big models the
00:05:06
um would call them networks of neurons but i call them uh
00:05:11
um statistical models and symbolic processing great strides have been
00:05:16
made in the last uh twenty years in the move towards
00:05:21
artificial intelligence even though unless it once again there is no intelligence in a high
00:05:29
this progress has to be seen in a very special context which
00:05:34
leads people to a call things into question this context started to develop
00:05:39
in the nineteen thirties with the invention of a advertising
00:05:46
uh this uh the uh early days of advertising
00:05:51
go back to nineteen thirties but a really became a
00:05:55
blatant in the nineteen fifties with the invention of self
00:05:57
service production consumption was multiplied by a factor of ten
00:06:03
and a waste was also increased by a factor of ten people put just and then throw things away straight away
00:06:09
increasingly with that advertising people began to consider humans as
00:06:14
part of the production tool and also as a consumption tool
00:06:19
and a few months almost became a product in themselves that's
00:06:23
what's happening with google and a face spoke many of us
00:06:27
don't realise this but we feel that the something wrong in
00:06:29
the system humans are increasingly becoming a product and are being exploited
00:06:38
if we reflect about capitalism and i am in
00:06:41
favour of capitalism but the face of capitalism has changed
00:06:45
prior to capitalism of people put used to meet needs
00:06:49
now we create needs in order to generate to drive production
00:06:57
so um uh production precedes the needs this is
00:07:02
serious i think that engineers are working on a i
00:07:05
have to learn to listen much more carefully to humans needs and to work on the basis of that
00:07:11
uh needs and that's what the telly molly foundation aims
00:07:15
to do rather than artificially creating a these uh needs
00:07:21
um the minivan appeared in sixties we knew the uh car
00:07:27
but uh no one said that they wanted to have something uh um
00:07:31
uh the the notes no one had a need for the minivan but um
00:07:36
someone just came up with that idea of something halfway between the car and to the uh
00:07:41
a track and the uh service uh it's the same thing for soups
00:07:50
on the other hand there's no intelligence in a high in forty years of work as a
00:07:56
an engineer in a i have i have never found the slightest signed of intelligence maybe the
00:08:02
uh next because will contradict me but i've never found a
00:08:07
single sign of intelligence in a a systems uh designed by engineers
00:08:14
but we should not denigrate the progress that's been made
00:08:20
uh we actually is working a i should not that great uh
00:08:24
the progress that's been made because it uh represents an
00:08:28
enormous potential progress and uh the possibility of contributing to society
00:08:35
with the technologies which are within our
00:08:38
reach we can greatly reduce parasitic tasks
00:08:44
in other words the tasks which are not worthy of a human being many people on the earth worker in
00:08:49
jobs which are not worthy of humans those jobs really have to be replaced as soon as possible by machines
00:08:57
and i know the to increase the cost of life of the people
00:09:01
who did those jobs previously and also to increase the quality of services
00:09:10
when i see that some people not only overestimate the difficulty tossed it also over
00:09:15
is over is the the the also underestimate the ability of the human a brain
00:09:23
humans are still um very capable of doing things which are computers it
00:09:27
cannot do one of the main characteristics of human beings is a empathy
00:09:32
yeah
00:09:35
welcome back to services numbers in later
00:09:41
one should really um the stress uh empathy and uh well when talking about providing services to humans
00:09:48
the technology that we have today really can contribute to improving the
00:09:53
flow of information and information is humanity it is the future of humanity
00:09:58
here i'm referring to going to refer to google sometimes google is going too far
00:10:05
i'm not talking about the top dressing but
00:10:08
when it comes to um making information available to everybody that uh it that's a good thing
00:10:17
but i think that sometimes when you've got the wrong a human uh the wrong economic model
00:10:22
humans should not be used as products they should be used as uses
00:10:27
the should be regarded as uses and the same thing applies to face spoke
00:10:32
connections between human beings are important that has to be improved
00:10:42
uh what was done to only twenty or thirty years ago on
00:10:45
a intelligent uh meeting rooms and a multilingual communication it's a good thing
00:10:51
but um now face because began to the bias the world
00:10:56
of media and the same applies to amazon amazon is it typical
00:11:00
uh example of how you can get rid
00:11:03
of 'em parser tickle tasks uh menial tasks uh
00:11:08
packaging things but now people are put on the trolleys and they have to
00:11:12
a gong get the um objects uh to be picked up and uh
00:11:17
they are once again being given a free menial tasks in fact
00:11:24
even you but uh which is a civilly criticising the initial mission of you but
00:11:30
uh enough cause uh uh the term used is uh you brought station
00:11:35
think that uh you bust initial uh approach was valid makes individual important uh
00:11:41
an individual doesn't have any uh uh doesn't have a job is not part was assuming what finds himself or herself useless
00:11:50
your graduation in the way is good because anyone can become a a little
00:11:53
entrepreneur but when they are excessive exported then the process is gone too far
00:12:03
there's no intelligence in a i just sound science which necessitates often not always
00:12:13
not always but often access to large databases which is sometimes very sensitive
00:12:20
so apart from the fact that humans are being used as a product there is this problem or access to databases
00:12:28
this is something which uh people cannot applies cannot express but they feel that something's wrong not really
00:12:33
uh other being used as products but in addition that data of being expected to own private data
00:12:41
and all this uh is taking place in an atmosphere of great to mistrust
00:12:49
so it's important to restore trust between humans and
00:12:53
a high and sensitive data clearly have to be protected
00:13:01
i've almost finished of stuck to my loaded speaking time
00:13:06
we have to share positive messages rather than negative messages that
00:13:11
you let me give you a few uh slides in english
00:13:17
all together instead of um disseminating fake news on the internet
00:13:24
i think we should work to increase trust between people who work on
00:13:31
a high and humans and i've provided a whole list of things
00:13:35
that we can do share posted messages uh provide open source information
00:13:43
create a content and uh access to information like pasta changes and quality of life
00:13:52
create a more efficient society increase transparency and encourage societal debates as what
00:13:58
we're doing here that's what the and tell him only a foundation does
00:14:03
what what you have to do is difficult but we have
00:14:05
to encourage societal debates and of course increased transparency and data privacy
00:14:16
uh_huh
00:14:20
addition to that
00:14:28
we have to create a safe zone uh for data this uh helps to build trust
00:14:37
in order to improve uh improve trust him data
00:14:42
uh_huh
00:14:45
so we have to know exactly what we're doing with data with postal data uh
00:14:53
so we have to share posted messages but not be too ambitious
00:15:00
where i'm making progress in a a vision we're making progress in the series and keep learning
00:15:08
conclusion
00:15:11
a i is full potential even though it doesn't have
00:15:14
any intelligence it's full potential but it should not be overestimated
00:15:22
please at the same time we should not underestimate the abilities of humans
00:15:31
or abilities will continue to be created what we will be able to achieve in the next few years we should never
00:15:37
uh turn away from the true missions intrude needs of humanity
00:15:43
we should never stray from that path and we should um
00:15:49
be wary of the false allegations that we find every day
00:15:53
on social media can be devastating sometimes scientists also played this game
00:16:01
the the every day they feel obliged to post things and some of those things are wrong we should
00:16:06
have confidence in transparency in use of data we should remain realistic and transparent in everything that we do
00:16:14
it's not make people believe in alchemy
00:16:21
probably just give us fire but uh that's not a
00:16:24
belief commits to into the vent it uh uh alchemy
00:16:30
oh
00:16:33
let's avoid that let's not believe in belgium alchemy
00:16:39
if to believe in science which was given to us but promiscuous and avoid getting caught up in alchemy it's remain
00:16:47
a realistic and continue to believe in the past it's of human beings
00:16:52
which go far beyond what we will ever be able to model that's
Conference Program
Welcome words
Aurélie Rosemberg, Fondation Dalle Molle
Sept. 11, 2021 · 4 p.m.
Artificial intelligence and quality of life
H. Bourlard, Idiap Research Institute
Sept. 11, 2021 · 4:30 p.m.
Artificial intelligence to think like humans
Melanie Mitchell, Professor at the Santa Fe Institute
Sept. 11, 2021 · 4:45 p.m.
Towards human-centered robotics
Sylvain Calinon, Research Director at the Idiap Research Institute
Sept. 11, 2021 · 5 p.m.
Supporting sustainable transitions around the world through water technology
Eric Valette, Director of AQUA4D
Sept. 11, 2021 · 5:15 p.m.
Biometric security
Sébastien Marcel, Research Director at the Idiap Research Institute
Sept. 11, 2021 · 5:30 p.m.
Compatibility with humans: AI and the problem of control
Stuart russel, Professor of Computer Science and Smith-Zadeh Professor of Engineering, University of California, Honorary Fellow of Berkeley and Wadham College at Oxford
Sept. 11, 2021 · 5:45 p.m.
Model subjectivity at the heart of consciousness to make robots more human
David Rudrauf, Associate professor at the University of Geneva, Director of the laboratory of the multimodal modeling of Emotion and Feeling
Sept. 11, 2021 · 6 p.m.
Round table
Panel
Sept. 11, 2021 · 6:15 p.m.
Recommended talks
Intelligence artificielle et qualité de la vie
Prof. Hervé Bourlard, Directeur de l'Idiap
Sept. 11, 2021 · 4:30 p.m.
Final pitch, team innovAItor
Lonneke, Idiap Research Institute
Aug. 26, 2021 · 3:16 p.m.
281 views