Click HERE to return to the contents page
Click HERE to return to the contents page
United Nations - Dr. David Hanson discusses about progress in artificial intelligence
Translation from English to Latin
Sophia Social Robot - for non-native English speakers
Published on Jan 5, 2018
A robot's "father" discusses progress in artificial intelligence and reveals a future that’s closer than you think.
00:00 We begin with Dr. David Hansen.
00:02 He is the founder and Chief Executive of Hanson Robotics,
00:04 which is the company responsible for Sophia. Dr. Hansen.
00:15 It's a it's a great honor to be here at the United Nations.
00:20 The ideals and hopes that the United Nations represents
00:26 for coming together to make a better future.
00:29 Drive me in my work with robots and I hope that we can connect with the
00:37 technologists of the world and great humanitarian and policy makers of the
00:43 world to decide how technology can benefit all.
00:50 The quest that we have at my company Hanson Robotics is to make machines
that are fundamentally human.
00:59 That have human attributes through bio-inspired engineering, so we take
the
01:06 mechanics of the Nano properties of human facial soft tissues.
01:14 We reverse-engineer that and we're able to produce robots that make more
01:18 natural-looking facial expressions with very low power and light weight.
01:22 That means that we can make mobile social robots. This is a new kind of
animation
01:27 art form and as an art form like animation has brought a lot of wonder
01:36 and delight to the world and been used for education for teaching children
01:40 about issues of the world. Robotics now can become a new kind of animation
art.
01:47 We can take that technology and combine it with artificial intelligence
and
01:53 It becomes a more natural interface for interacting with artificial
intelligence.
01:57 Making artificial intelligence more emotionally accessible to people.
02:03 We also want artificial intelligence to really understand what it means to
be human?
02:09 So having a natural interaction with people allows machines
02:14 to form a relationship with people, to learn and grow with people, not
separate from people
02:20 By making AI that grows up among people.
02:24 There is the possibility that when and if AI artificial intelligence and
robotics becomes alive.
02:34 That it really cares about us. There is a tantalizing prospect that
02:39 we could transition from mirror bio-inspired engineering to actual
02:44 bioengineering through our robotics and artificial intelligence
technology.
02:48 There's a revolution at work today in the field of biological engineering,
02:56 protein engineering, genomics the engineering of new life-forms through
crisper.
03:03 This is real we have just begun to see the implications of this work it
03:08 will change the world. We're also understanding the human mind
03:13 better than we ever have there are still deep mysteries we don't have a
full
03:18 model of human intelligence of consciousness of the workings of the brain.
03:23 However there was a revolution in connect domains the mapping of the human
03:28 mind and it's information processing systems and how that interoperates
with
03:33 the systems biology of the human body of proteins of hormones of the
genome.
03:42 So as these scientists come together we are going to be able to make
machines that
03:48 are more like humans and biological machines that are inspired by the very
03:54 depths of biology. This means that machines across the
03:58 board are coming to life. When will they've come to life as
04:02 complete organisms we don't know that. However when they do I want to make
sure
04:08 that they care about us that they embody the best of human values and not
the worst.
04:15 This is the quest we've got to find a way to do better to be the best.
04:19 That we can be and to make machines that reflect the best of who we can be
as well.
04:26 So at Hanson Robotics when we say we bring robots to life.
04:32 Next slide please. This means that we are humanizing robots
04:38 as a new kind of animation art form. We are also providing a natural
interface
04:43 for artificial intelligence and striving to make machines that are truly
04:48 generally intelligent, adaptive in the way of organisms and do so in such
a way
04:54 that they really care about the world can understand the consequences of
their
04:59 actions and strive to make the world a better place for people and all
living beings.
05:06 Next slide. So we are developing a series of characters as experiments as
05:13 science as art and as products. So we have made a small Einstein robot,
05:22 we have made this Sophia which you see here today and we have Han who you
see on
05:31 this slide and they're going out as effectively ambassadors of this vision
05:36 that technology can really help to make people's lives better. Next slide.
05:41 The professor Einstein is teaching children's science and playing brain
05:45 games with children teaching them about relativity and this is a first
step.
05:51 Now this kind of social robot can become a general purpose platform for
how people
05:55 interact with technology. Voice assistants are very common in the world
05:59 and they're using a natural kind of interface. However human to human
06:03 interactions are only partially voice-based we are also very rich visual
06:08 social creatures. We read each other's facial expressions and body
language
06:12 that's why we love animation and why we love characters in films that like
06:17 actors, that are acting with their faces. So we're bringing this to the
interface
06:22 to future AI agents and that's what. Professor Einstein represents. Next
slide.
06:30 We are also developing open-source platforms for kids to
06:34 develop their own characters and a good example of this we call Dr. Roboto.
06:38 This is a completely open platform reprogrammable hackable and you can
06:43 develop it with new AI. Next slide. We're developing things with Disney
we're
06:49 doing outreach with a broader community as well. Next slide please.
06:56 Next slide please. So with Sofia Quest we're putting out an
07:02 open-source general purpose platform for AI artificial general
intelligence and
07:07 hardware development. We have made open-source arms that you see here that
07:11 other people can develop they can download and improve these.
07:15 Open-source designs physical designs for the head and open source software
for researching
07:20 social robots, natural language understanding, natural language
interactions.
07:26 And when she encounters people what's interesting is that people
07:30 open up they react warmly to Sofia they afterwards we often get messages
about
07:37 how she's doing and what's she doing in her life. So this is very
interesting it
07:42 means that we can form a kind of relationship with it our technology that
07:47 may not have been possible before. Next slide. Now most robots in the
world are
07:54 not human we intentionally make them not human we treat them as tools.
08:00 Our AI algorithms often are developed to run on server farms completely
out of sight.
08:08 Now what happens when these kinds of machines awaken if they become truly
08:15 alive and start to care about the world? Well their alien they're not
designed
08:21 necessarily for social interaction however computer animation has been a
08:27 very successful industry making animated characters that people care about
people
08:33 fall in love with these kinds of characters not just for movies but for
08:36 video games as well. So in the history of making these kinds of humanoid
robots
08:50 if you'll click enter one more time then these kinds of click on that
Einstein
08:59 face up on the upper left if you will.
09:03 These kinds of there we go robots can show a full range of natural
09:09 facial expressions they can mount on walking robot bodies like the human
robot
09:14 which subsequently run won the dark Robotics Challenge. Which means that
you
09:20 can see these robots prefer performing actual search-and-rescue cowork
applications.
09:27 General service applications and there are now a full
09:30 animated interface. Next slide please.
09:35 Please keep going. Next slide. Where this fits in is the idea of whole
organism
09:42 cognition this is I believe we're the next revolution and artificial
09:45 intelligence work will come from. So we've engineered this whole organism
09:50 cognition Network. Next slide. Into a block chain safe and secure hyper
data
09:59 compliant network that we call singularity net. And singularity net is
10:05 designed specifically for AI to understand the consequences and maximize
net benefits.
10:11 The goal is a new economic measure called maximum net benefit.
10:15 For all profits, human rights preserving the ecosystems and there is this
possibility
10:26 that machines may become alive in our lifetime you may this is derived
from a
10:30 variety of economic statistics. That show and and technology performance.
Next slide.
10:37 If these machines like we see in this sort of research that looked like
10:42 with Boston Dynamics and ASIMO. If they become alive they've got to care
about
10:49 us we need to make sure that they're safe. So processes like singularity
net.
10:54 Next slide. The goal of making machines that are truly ethical instead of
just
11:00 enslaved machines that might come alive. This is where this goal of
maximized net benefits.
11:08 Becomes really essential we've got to wire our machines for safety now
11:13 and for love at this time not when they become alive. Next slide.
11:20 So thank you very much.
11:22
* Click the link to watch the video and select your native or favorite
Vocabulary and Auto-translate.
11:25
* Please Add your own subtitles or correct existing (Settings -
Subtitles/CC - Add).
11:29 Thank you Dr. Hansen.
