Towards Artificial General Intelligence | Oriol Vinyals | TEDxImperialCollege

I have one thank you so I hope most of

I have one thank you so I hope most of you have heard it’s the term artificial

you have heard it’s the term artificial

you have heard it’s the term artificial intelligence or artificial general

intelligence or artificial general

intelligence or artificial general intelligence the sort of textbook

intelligence the sort of textbook

intelligence the sort of textbook definition would be sort of the study of

definition would be sort of the study of

definition would be sort of the study of building agent as we call them to solve

building agent as we call them to solve

building agent as we call them to solve a variety a large variety of very

a variety a large variety of very

a variety a large variety of very complex tasks before without needing to

complex tasks before without needing to

complex tasks before without needing to be reprogrammed so these is a very long

be reprogrammed so these is a very long

be reprogrammed so these is a very long quest which we have taken on and there’s

quest which we have taken on and there’s

quest which we have taken on and there’s still a long way to go for us to sort of

still a long way to go for us to sort of

still a long way to go for us to sort of see it done but crucially it’s already

see it done but crucially it’s already

see it done but crucially it’s already shown a lot of advancements and a lot of

shown a lot of advancements and a lot of

shown a lot of advancements and a lot of progress that we already seen our day

progress that we already seen our day

progress that we already seen our day today but this thought I actually want

today but this thought I actually want

today but this thought I actually want to explain how we the researchers are

to explain how we the researchers are

to explain how we the researchers are actually trying to achieve this goal

actually trying to achieve this goal

actually trying to achieve this goal so not only like define it but actually

so not only like define it but actually

so not only like define it but actually how are we chuckling nowadays building

how are we chuckling nowadays building

how are we chuckling nowadays building idea so imagine for a second that I

idea so imagine for a second that I

idea so imagine for a second that I asked you to be sort of a bit of an

asked you to be sort of a bit of an

asked you to be sort of a bit of an engineer perhaps and do something that

engineer perhaps and do something that

engineer perhaps and do something that to us looks extremely trivial we do this

to us looks extremely trivial we do this

to us looks extremely trivial we do this every day so there’s a robot and I ask

every day so there’s a robot and I ask

every day so there’s a robot and I ask you okay you build an algorithm a

you okay you build an algorithm a

you okay you build an algorithm a sequence of instructions to make this

sequence of instructions to make this

sequence of instructions to make this robot open the door now it is very

robot open the door now it is very

robot open the door now it is very common to sort of break out this problem

common to sort of break out this problem

common to sort of break out this problem into several pieces and what you could

into several pieces and what you could

into several pieces and what you could do is sort of say do that right so to

do is sort of say do that right so to

do is sort of say do that right so to open up find the door find where the

open up find the door find where the

open up find the door find where the handle of the door is grab it push it

handle of the door is grab it push it

handle of the door is grab it push it down and open it I in turns out out that

down and open it I in turns out out that

down and open it I in turns out out that if you actually strictly speaking do

if you actually strictly speaking do

if you actually strictly speaking do this sort of rule-based and hierarchical

this sort of rule-based and hierarchical

this sort of rule-based and hierarchical way to solve a complex problem breaking

way to solve a complex problem breaking

way to solve a complex problem breaking it down into smaller not so complex

it down into smaller not so complex

it down into smaller not so complex problems you end up with something that

problems you end up with something that

problems you end up with something that perhaps is not so optimal

perhaps is not so optimal

perhaps is not so optimal so if you simply try to what’s going to

so if you simply try to what’s going to

so if you simply try to what’s going to happen if somebody open you are not

happen if somebody open you are not

happen if somebody open you are not accounting for these the terrain is not

accounting for these the terrain is not

accounting for these the terrain is not as you know now as you have hope and the

as you know now as you have hope and the

as you know now as you have hope and the key concept here is that your algorithm

key concept here is that your algorithm

key concept here is that your algorithm will not generalize to the complexity of

will not generalize to the complexity of

will not generalize to the complexity of the world this by the way is from the

the world this by the way is from the

the world this by the way is from the darker competition in 2015

darker competition in 2015

darker competition in 2015 these programs are not really as simple

these programs are not really as simple

these programs are not really as simple as the four lines I showed you and it

as the four lines I showed you and it

as the four lines I showed you and it really shows that trying to build

really shows that trying to build

really shows that trying to build algorithms to these like rule-based

algorithms to these like rule-based

algorithms to these like rule-based chain of rules

chain of rules

chain of rules Jorge really sort of is bound to failure

Jorge really sort of is bound to failure

Jorge really sort of is bound to failure in mistakes of course these were like

in mistakes of course these were like

in mistakes of course these were like the funny ones let’s try to go to the

the funny ones let’s try to go to the

the funny ones let’s try to go to the next line yeah so here is how you could

next line yeah so here is how you could

next line yeah so here is how you could do this perhaps in a slightly different

do this perhaps in a slightly different

do this perhaps in a slightly different way I call this the learning approach

way I call this the learning approach

way I call this the learning approach and this is studying a lot knowledge so

and this is studying a lot knowledge so

and this is studying a lot knowledge so take the problem of finding the door

take the problem of finding the door

take the problem of finding the door right how would you program that right

right how would you program that right

right how would you program that right so if you you were maybe you have these

so if you you were maybe you have these

so if you you were maybe you have these pixels you try to find these rectangular

pixels you try to find these rectangular

pixels you try to find these rectangular objects that look like doors perhaps and

objects that look like doors perhaps and

objects that look like doors perhaps and as you saw there it’s not gonna channel

as you saw there it’s not gonna channel

as you saw there it’s not gonna channel eyes to any conditions you may have not

eyes to any conditions you may have not

eyes to any conditions you may have not imagined and then you have to try on

imagined and then you have to try on

imagined and then you have to try on error iterate and this really doesn’t

error iterate and this really doesn’t

error iterate and this really doesn’t scale because you’re just now opening

scale because you’re just now opening

scale because you’re just now opening doors and how about doing anything else

doors and how about doing anything else

doors and how about doing anything else so the pattern here is I’m gonna instead

so the pattern here is I’m gonna instead

so the pattern here is I’m gonna instead try to get a lot of pictures of doors

try to get a lot of pictures of doors

try to get a lot of pictures of doors and I have human laborers labeling where

and I have human laborers labeling where

and I have human laborers labeling where the doors are and what I gonna do is I’m

the doors are and what I gonna do is I’m

the doors are and what I gonna do is I’m gonna learn what a lot door looks like

gonna learn what a lot door looks like

gonna learn what a lot door looks like instead of programming in terms of rules

instead of programming in terms of rules

instead of programming in terms of rules and it is very very important so

and it is very very important so

and it is very very important so unfortunately in computer vision people

unfortunately in computer vision people

unfortunately in computer vision people care about other objects knock doors and

care about other objects knock doors and

care about other objects knock doors and people generally don’t take pictures of

people generally don’t take pictures of

people generally don’t take pictures of after something I showed you the example

after something I showed you the example

after something I showed you the example of finding actually cats and dogs which

of finding actually cats and dogs which

of finding actually cats and dogs which are very popular objects and in the

are very popular objects and in the

are very popular objects and in the research community so how are we

research community so how are we

research community so how are we actually gonna do that so the first step

actually gonna do that so the first step

actually gonna do that so the first step to finding that cat whether an image

to finding that cat whether an image

to finding that cat whether an image contains let’s say a cat or a dog is to

contains let’s say a cat or a dog is to

contains let’s say a cat or a dog is to build a big database of pictures that

build a big database of pictures that

build a big database of pictures that contain iron cats or dogs and then

contain iron cats or dogs and then

contain iron cats or dogs and then literally as humans to label thousands

literally as humans to label thousands

literally as humans to label thousands or tens of thousands

or tens of thousands

or tens of thousands examples of images like this one which

examples of images like this one which

examples of images like this one which actually is a dog although this one is

actually is a dog although this one is

actually is a dog although this one is pretty tricky I would say so how would

pretty tricky I would say so how would

pretty tricky I would say so how would you even you know code these and go

you even you know code these and go

you even you know code these and go through the pictures and so on so the

through the pictures and so on so the

through the pictures and so on so the first piece is get these database of

first piece is get these database of

first piece is get these database of labeled examples once you have these the

labeled examples once you have these the

labeled examples once you have these the second step is allow for a generic

second step is allow for a generic

second step is allow for a generic program to be able to learn the

program to be able to learn the

program to be able to learn the characteristics or the patterns on these

characteristics or the patterns on these

characteristics or the patterns on these database and that’s precisely what deep

database and that’s precisely what deep

database and that’s precisely what deep learning and in particular give neural

learning and in particular give neural

learning and in particular give neural nets are enabling us to do so here you

nets are enabling us to do so here you

nets are enabling us to do so here you see a big neural network that actually

see a big neural network that actually

see a big neural network that actually processes the image that were pixels and

processes the image that were pixels and

processes the image that were pixels and activates certain neurons in this

activates certain neurons in this

activates certain neurons in this architecture that could well be

architecture that could well be

architecture that could well be for instance detecting the size of the

for instance detecting the size of the

for instance detecting the size of the animal the floor enos whether it has pig

animal the floor enos whether it has pig

animal the floor enos whether it has pig ears and so on and crucially these

ears and so on and crucially these

ears and so on and crucially these features that these neural network is

features that these neural network is

features that these neural network is activating on this image and not

activating on this image and not

activating on this image and not prepared by us these are fully learned

prepared by us these are fully learned

prepared by us these are fully learned from these data backs so again we have

from these data backs so again we have

from these data backs so again we have the database we have this model that is

the database we have this model that is

the database we have this model that is powerful enough to represent all sorts

powerful enough to represent all sorts

powerful enough to represent all sorts of interesting features that eventually

of interesting features that eventually

of interesting features that eventually lead to an output that is the

lead to an output that is the

lead to an output that is the probability that the image contains

probability that the image contains

probability that the image contains attack or dock and what you essentially

attack or dock and what you essentially

attack or dock and what you essentially do is feed one image at a time to this

do is feed one image at a time to this

do is feed one image at a time to this network of the training database and ask

network of the training database and ask

network of the training database and ask it to predict the correct label all the

it to predict the correct label all the

it to predict the correct label all the time eventually all these sort of

time eventually all these sort of

time eventually all these sort of connections settle onto a configuration

connections settle onto a configuration

connections settle onto a configuration that works well and generalizes beyond

that works well and generalizes beyond

that works well and generalizes beyond the images that means being trainer and

the images that means being trainer and

the images that means being trainer and internally so well that in fact the

internally so well that in fact the

internally so well that in fact the billions of pictures every single day

billions of pictures every single day

billions of pictures every single day are being analyzed through this kind of

are being analyzed through this kind of

are being analyzed through this kind of algorithm so the next and one of the

algorithm so the next and one of the

algorithm so the next and one of the sort of breakthroughs in 2014 which me

sort of breakthroughs in 2014 which me

sort of breakthroughs in 2014 which me incorporators and other groups are sort

incorporators and other groups are sort

incorporators and other groups are sort of discovered is not only this idea of

of discovered is not only this idea of

of discovered is not only this idea of taking an image and producing a

taking an image and producing a

taking an image and producing a probability but also taking a sort of

probability but also taking a sort of

probability but also taking a sort of sequence of symbols and mapping it

sequence of symbols and mapping it

sequence of symbols and mapping it through learning examples to another

through learning examples to another

through learning examples to another sequence of symbols and this is a very

sequence of symbols and this is a very

sequence of symbols and this is a very powerful

powerful

powerful because now we can also learn for

because now we can also learn for

because now we can also learn for instance to do machine translation so

instance to do machine translation so

instance to do machine translation so here the sequence the input sequence

here the sequence the input sequence

here the sequence the input sequence would be the Chinese factors and the

would be the Chinese factors and the

would be the Chinese factors and the output would be English works and

output would be English works and

output would be English works and crucially you learn these without any

crucially you learn these without any

crucially you learn these without any knowledge of a language or which words

knowledge of a language or which words

knowledge of a language or which words correspond to which words all you need

correspond to which words all you need

correspond to which words all you need is a large database of pairs of

is a large database of pairs of

is a large database of pairs of sentences in Chinese and in English

sentences in Chinese and in English

sentences in Chinese and in English translations of each other and from a

translations of each other and from a

translations of each other and from a few of these pairs of examples you

few of these pairs of examples you

few of these pairs of examples you essentially achieve a system that

essentially achieve a system that

essentially achieve a system that actually generalizes AI improves upon

actually generalizes AI improves upon

actually generalizes AI improves upon the real estate of the art so in fact

the real estate of the art so in fact

the real estate of the art so in fact Google Translate has been upgraded

Google Translate has been upgraded

Google Translate has been upgraded recently to use these sort of neural

recently to use these sort of neural

recently to use these sort of neural networks approaches that essentially

networks approaches that essentially

networks approaches that essentially reach the gap between the previous best

reach the gap between the previous best

reach the gap between the previous best systems and human quality translations

systems and human quality translations

systems and human quality translations by F by 50% essentially so these again

by F by 50% essentially so these again

by F by 50% essentially so these again is already used and it’s sort of already

is already used and it’s sort of already

is already used and it’s sort of already out there for everyone to benefit from

out there for everyone to benefit from

out there for everyone to benefit from another such example that is sort of

another such example that is sort of

another such example that is sort of interesting is generating speech so we

interesting is generating speech so we

interesting is generating speech so we actually do understand how speech is

actually do understand how speech is

actually do understand how speech is generated right there is a vocal tract

generated right there is a vocal tract

generated right there is a vocal tract which essentially is a bunch of tubes

which essentially is a bunch of tubes

which essentially is a bunch of tubes that check changing shape and you could

that check changing shape and you could

that check changing shape and you could actually model this and this is one of

actually model this and this is one of

actually model this and this is one of these parametric approaches that you see

these parametric approaches that you see

these parametric approaches that you see there on the left but you can also do is

there on the left but you can also do is

there on the left but you can also do is just record this acoustic the audio

just record this acoustic the audio

just record this acoustic the audio files essentially from many people which

files essentially from many people which

files essentially from many people which will do very complex from fire there’s

will do very complex from fire there’s

will do very complex from fire there’s like tens of thousands of points per

like tens of thousands of points per

like tens of thousands of points per second of speech but again you just let

second of speech but again you just let

second of speech but again you just let a neural network figure out the sort of

a neural network figure out the sort of

a neural network figure out the sort of patterns in this signal and what you get

patterns in this signal and what you get

patterns in this signal and what you get end up with is a very good natural

end up with is a very good natural

end up with is a very good natural sounding speech system that again beats

sounding speech system that again beats

sounding speech system that again beats the gap between the previous best and

the gap between the previous best and

the gap between the previous best and human speech by about 50% factor so this

human speech by about 50% factor so this

human speech by about 50% factor so this is some like drastic already sort of

is some like drastic already sort of

is some like drastic already sort of well-developed ivy

well-developed ivy

well-developed ivy that supervised learning enable us to do

that supervised learning enable us to do

that supervised learning enable us to do and it’s all about this idea of

and it’s all about this idea of

and it’s all about this idea of imitation learning you imitate humans

imitation learning you imitate humans

imitation learning you imitate humans through these powerful models deep

through these powerful models deep

through these powerful models deep neural Nets so now there is another

neural Nets so now there is another

neural Nets so now there is another paradigm I wanted to describe which is

paradigm I wanted to describe which is

paradigm I wanted to describe which is what if we cannot get a large database

what if we cannot get a large database

what if we cannot get a large database of these examples or even better we want

of these examples or even better we want

of these examples or even better we want to actually go beyond human capabilities

to actually go beyond human capabilities

to actually go beyond human capabilities right if we actually want to beat humans

right if we actually want to beat humans

right if we actually want to beat humans if we take that we are kind of bounded

if we take that we are kind of bounded

if we take that we are kind of bounded by how good they are and this is the

by how good they are and this is the

by how good they are and this is the reinforcement learning paradigm and for

reinforcement learning paradigm and for

reinforcement learning paradigm and for instance in the game of gold this was

instance in the game of gold this was

instance in the game of gold this was heavily used to for the recent result

heavily used to for the recent result

heavily used to for the recent result that I forgot chief de la fere so it is

that I forgot chief de la fere so it is

that I forgot chief de la fere so it is extremely simple how this works it’s

extremely simple how this works it’s

extremely simple how this works it’s almost like obvious the first step is we

almost like obvious the first step is we

almost like obvious the first step is we have an agent which is in this case the

have an agent which is in this case the

have an agent which is in this case the top layer and this agent lives in an

top layer and this agent lives in an

top layer and this agent lives in an environment so the imagine would be the

environment so the imagine would be the

environment so the imagine would be the dog or or in the robot case the world

dog or or in the robot case the world

dog or or in the robot case the world and this agent has a dog for instance it

and this agent has a dog for instance it

and this agent has a dog for instance it wants to win at the game of God and the

wants to win at the game of God and the

wants to win at the game of God and the way this works is quite simple the agent

way this works is quite simple the agent

way this works is quite simple the agent observes the environment so it observes

observes the environment so it observes

observes the environment so it observes let’s say the goal board or the door or

let’s say the goal board or the door or

let’s say the goal board or the door or what one not and it can also act upon

what one not and it can also act upon

what one not and it can also act upon the embark it can for instance play a

the embark it can for instance play a

the embark it can for instance play a piece of the port that would be one

piece of the port that would be one

piece of the port that would be one action right I play in position four

action right I play in position four

action right I play in position four coma fight for instance in the in that

coma fight for instance in the in that

coma fight for instance in the in that sort of game and eventually six I cling

sort of game and eventually six I cling

sort of game and eventually six I cling to these the game finishes and there my

to these the game finishes and there my

to these the game finishes and there my environment as creation whirring won the

environment as creation whirring won the

environment as creation whirring won the game or not and after it’s been taught

game or not and after it’s been taught

game or not and after it’s been taught whether or not it updates its ways to

whether or not it updates its ways to

whether or not it updates its ways to ensure that essentially the the reward

ensure that essentially the the reward

ensure that essentially the the reward that it gets the probability of winning

that it gets the probability of winning

that it gets the probability of winning is maximized

is maximized

is maximized so this agent is very very similar to

so this agent is very very similar to

so this agent is very very similar to the neural network that was classifying

the neural network that was classifying

the neural network that was classifying cancer dogs but in this case it takes as

cancer dogs but in this case it takes as

cancer dogs but in this case it takes as the inputs these observations so it

the inputs these observations so it

the inputs these observations so it would be a goal board

would be a goal board

would be a goal board video games cleaner or not and the

video games cleaner or not and the

video games cleaner or not and the outputs are simply the probability of

outputs are simply the probability of

outputs are simply the probability of each actions or which joystick position

each actions or which joystick position

each actions or which joystick position you want to play and through this

you want to play and through this

you want to play and through this process you can achieve superhuman

process you can achieve superhuman

process you can achieve superhuman performance in playing at I’ll go and so

performance in playing at I’ll go and so

performance in playing at I’ll go and so on so very crucially it is most of these

on so very crucially it is most of these

on so very crucially it is most of these algorithms have been developed

algorithms have been developed

algorithms have been developed developing games um we just kind of

developing games um we just kind of

developing games um we just kind of funny because as a kid I used to play

funny because as a kid I used to play

funny because as a kid I used to play video games and now I sort of work on

video games and now I sort of work on

video games and now I sort of work on making algorithms for video games I

making algorithms for video games I

making algorithms for video games I don’t get to play but despite these

don’t get to play but despite these

don’t get to play but despite these profound you know realizations what you

profound you know realizations what you

profound you know realizations what you actually get to do when you do through

actually get to do when you do through

actually get to do when you do through these two video games is being able to

these two video games is being able to

these two video games is being able to scale because as you can imagine this

scale because as you can imagine this

scale because as you can imagine this circle of observations and actions

circle of observations and actions

circle of observations and actions initially it’s not very good so in other

initially it’s not very good so in other

initially it’s not very good so in other words if you act randomly you would not

words if you act randomly you would not

words if you act randomly you would not gonna get the works very often so you

gonna get the works very often so you

gonna get the works very often so you have to play a lot in order to achieve

have to play a lot in order to achieve

have to play a lot in order to achieve good performance but if you go out to

good performance but if you go out to

good performance but if you go out to the real world

the real world

the real world things will be very slow if you are to

things will be very slow if you are to

things will be very slow if you are to build 10,000 products that you would

build 10,000 products that you would

build 10,000 products that you would crash and eventually you learn how to do

crash and eventually you learn how to do

crash and eventually you learn how to do this that wouldn’t scare everyone so

this that wouldn’t scare everyone so

this that wouldn’t scare everyone so video games are very important because

video games are very important because

video games are very important because you can’t run them in parallel in a

you can’t run them in parallel in a

you can’t run them in parallel in a cluster and if you lose it’s not a big

cluster and if you lose it’s not a big

cluster and if you lose it’s not a big deal but eventually you are testing sort

deal but eventually you are testing sort

deal but eventually you are testing sort of ideas about intelligence through this

of ideas about intelligence through this

of ideas about intelligence through this video guys which were designed by making

video guys which were designed by making

video guys which were designed by making essentially humans you know imagine and

essentially humans you know imagine and

essentially humans you know imagine and plan and so on so this is our sort of

plan and so on so this is our sort of

plan and so on so this is our sort of research platform it has been so far

research platform it has been so far

research platform it has been so far until recently Alex it’s gonna be in the

until recently Alex it’s gonna be in the

until recently Alex it’s gonna be in the future as well but what’s the problem

future as well but what’s the problem

future as well but what’s the problem well the environment if the environment

well the environment if the environment

well the environment if the environment is very simple we will learn by simple

is very simple we will learn by simple

is very simple we will learn by simple agents so the environment and the agents

agents so the environment and the agents

agents so the environment and the agents go side by side and as you can imagine

go side by side and as you can imagine

go side by side and as you can imagine the

the

the is extremely complex so there are like

is extremely complex so there are like

is extremely complex so there are like you know predators out there that might

you know predators out there that might

you know predators out there that might want to eat you you need to energy so

want to eat you you need to energy so

want to eat you you need to energy so you might want to eat plants or other

you might want to eat plants or other

you might want to eat plants or other other human being not human beings

other human being not human beings

other human being not human beings beings and and it’s very diverse and

beings and and it’s very diverse and

beings and and it’s very diverse and very complex right so it seems like

very complex right so it seems like

very complex right so it seems like creating environments and creating

creating environments and creating

creating environments and creating agents goes both hand by hand and he

agents goes both hand by hand and he

agents goes both hand by hand and he didn’t need that so one of the the ideas

didn’t need that so one of the the ideas

didn’t need that so one of the the ideas here is if we are simulating the the

here is if we are simulating the the

here is if we are simulating the the natural world and we want for instance

natural world and we want for instance

natural world and we want for instance to get this guy to like survive the

to get this guy to like survive the

to get this guy to like survive the winter we we need to simulate snow and

winter we we need to simulate snow and

winter we we need to simulate snow and food and so on and that is a hard

food and so on and that is a hard

food and so on and that is a hard problem by itself as well however you

problem by itself as well however you

problem by itself as well however you know what you can actually try to do is

know what you can actually try to do is

know what you can actually try to do is well we researchers not only build

well we researchers not only build

well we researchers not only build agents but we are actually also seeing

agents but we are actually also seeing

agents but we are actually also seeing about the environments that will make

about the environments that will make

about the environments that will make agents advance and amongst many new

agents advance and amongst many new

agents advance and amongst many new environments that people create last

environments that people create last

environments that people create last year we created this environment which

year we created this environment which

year we created this environment which is called the mine lab which looks like

is called the mine lab which looks like

is called the mine lab which looks like three dimensional you know researchers

three dimensional you know researchers

three dimensional you know researchers can set all sorts of tasks to test to

can set all sorts of tasks to test to

can set all sorts of tasks to test to test ideas about planning memory

test ideas about planning memory

test ideas about planning memory imagination and so on and I’ll show you

imagination and so on and I’ll show you

imagination and so on and I’ll show you just a simple example where an agent

just a simple example where an agent

just a simple example where an agent will essentially go around and try to

will essentially go around and try to

will essentially go around and try to eat apples but do not eat lemons this is

eat apples but do not eat lemons this is

eat apples but do not eat lemons this is another video where there’s actually

another video where there’s actually

another video where there’s actually like more naturalistic looking video and

like more naturalistic looking video and

like more naturalistic looking video and there’s an agent out there that is sort

there’s an agent out there that is sort

there’s an agent out there that is sort of navigating these complex works

of navigating these complex works

of navigating these complex works figuring out that apples might appear

figuring out that apples might appear

figuring out that apples might appear again because they grow and so on and so

again because they grow and so on and so

again because they grow and so on and so forth so these environments are becoming

forth so these environments are becoming

forth so these environments are becoming more complex and as a result our agents

more complex and as a result our agents

more complex and as a result our agents will also of course learn better they’re

will also of course learn better they’re

will also of course learn better they’re now a sort of a final note it is kind of

now a sort of a final note it is kind of

now a sort of a final note it is kind of obvious that we need AI it’s very

obvious that we need AI it’s very

obvious that we need AI it’s very helpful to you know be able to have

helpful to you know be able to have

helpful to you know be able to have translation systems you might have a

translation systems you might have a

translation systems you might have a person

person

person system in your phone like helping you

system in your phone like helping you

system in your phone like helping you organize streets and so on so forth but

organize streets and so on so forth but

organize streets and so on so forth but it also seems a friend that an AI needs

it also seems a friend that an AI needs

it also seems a friend that an AI needs nature we need to encode these complex

nature we need to encode these complex

nature we need to encode these complex rules for AI to be useful to us

rules for AI to be useful to us

rules for AI to be useful to us otherwise we are just bound by the

otherwise we are just bound by the

otherwise we are just bound by the complexity of the environment in which

complexity of the environment in which

complexity of the environment in which agents living and so we done note them

agents living and so we done note them

agents living and so we done note them I’d like to just thank you for your

I’d like to just thank you for your

I’d like to just thank you for your attention I’ll be happy to chat with you

attention I’ll be happy to chat with you

attention I’ll be happy to chat with you outside Thanks

outside Thanks

outside Thanks [Music]