2020 WIPO-Korea

Summer School

AI and IP
10 July 2020

Gyudong HAN
Director of Information Management Division

KIPO
 Why do we need to know about AI?
Deepfake AI vs Bias

* deepfake(1:11) : https://www.youtube.com/watch?v=cQ54GDm1eL0

* Google news search 2

 Why dotowe
We need need
know theto knowof
reality about
AI AI?
AI, hype?
AI vs Job Starsky Robotics CEO Stefan Seltz-Axmacher,
Shutting down the company, 2020.3.20.
“There aren’t many startups
that can survive 10 years without shipping”

* Google news search

* The End of Starsky Robotics, https://medium.com/starsky-robotics-blog/the-end-of-starsky-robotics-acb8a6a8a5f5 3
 Table of Contents

 What is Artificial Intelligence?

 Artificial Intelligence(AI)

 Machine Learning(ML)

 Artificial Neural Network(ANN)

 Deep Learning(DL)

 AI-related Intellectual Property(IP) issues

 AI and IP automation

 Patent policy issues

 Copyright policy issues

4
01 What is AI?
What is the first thing that comes to mind
when you hear the word “AI”?

A big shock AI movies

to Koreans
Terminator, Avengers
Her, Transcendence
…

AI In
daily life

* http://www.bbc.com/news/technology-35785875
* Google search 5
 That’s partially because the
AI, difficult to define notion has changed over time

John McCarthy coined the term

“Artificial Intelligence”
which he would define as
“the science and engineering
of making intelligent machines”.
What is intelligent machine?

 Thinking Humanly: new effort to make computers think

 Acting Humanly: the art of creating machines that perform
functions that require intelligence
 Thinking Rationally: the study of the computations that
make it possible to perceive, reason and act
 Acting Rationally: computational intelligence is the study
of the design of intelligent agents
A rational agent is one that acts so as to achieve the best outcome or,
when there is uncertainty, the best expected outcome.
* https://www.artificial-solutions.com/blog/homage-to-john-mccarthy-the-father-of-artificial-intelligence
* “Artificial Intelligence A Modern Approach” 3rd Edition 1~2page, Stuart Russell, Peter Norvig, Prentice Hall 6
 Difficult to define because
 AI effect: Concept of AI has changed over time
 Once something becomes commonplace, it's
demystified. And it doesn't feel like the magical
intelligence that we see in humans. (ex. search engine, OCR)
The concept varies
 AI has two categories from person to person
 AI (Artificial Intelligence): Try to imitate human brain or
express human thoughts in algorithms
 IA (Intelligence Augmentation): It’s impossible for a
computer to imitate human intelligence and it should
be focused on technologies that help strengthen
human intelligence. (ex. IBM Watson, Google)
 Today, AI = IA + AI
* WHO AM I? https://www.youtube.com/watch?v=UHwVyplU3Pg, edited partially
* AI effect http://www.businessinsider.com/misconception-artificial-intelligene-2015-10 / AI vs IA https://iac.nia.or.kr/board_files/70/download 7
 [Book title: AI and
Deep Learning]
Levels of AI

 [Level 1] Simple control program

 Advertise a simple control program of a machine
as “AI applied” (washing machine, vacuum cleaner)

 [Level 2] Complex control program

 The relationship of input and output is large and complex
(traditional puzzle game, expert system with knowledge base)

 [Level 3] AI with Machine Learning

 The performance of level 2 machine can be improved to level 3
by machine learning . But feature extraction is done by human
 [Level 4] AI with Deep Learning
 The input features are extracted not by human but by machines

* “인공지능과 딥러닝”, 마쓰오 유타카, 동아 엠엔비 [Book: “AI and deep learning”, Yutaka Matsuo, Donga M&B] 8
 Areas of AI
What’s required for a machine to be intelligent?

* http://futurearchitectureplatform.org/news/28/ai-architecture-intelligence/

Predictive analytics : statistical techniques from predictive modelling, machine learning, and data mining
that analyze current and historical facts to make predictions about future or otherwise unknown events
* https://en.wikipedia.org/wiki/Predictive_analytics 9
 Relations between AI, ML and DL

Artificial Neural Network

AI other than ML

Search, Inference,
Expert system etc

* http://blogs.nvidia.co.kr/2016/08/03/difference_ai_learning_machinelearning/
10
 AI, a marketing strategy?

Machine Learning has certainly been seized as an opportunity by

marketers. After AI has been around for so long, it’s possible that it
started to be seen as something that’s in some way “old hat” even
before its potential has ever truly been achieved. There have been a
few false starts along the road to the “AI revolution”, and the term
Machine Learning certainly gives marketers something new, shiny
and, importantly, firmly grounded in the here-and-now, to offer.
* What Is The Difference Between Artificial Intelligence And Machine Learning?, Forbes, Dec 6, 2016
https://www.forbes.com/sites/bernardmarr/2016/12/06/what-is-the-difference-between-artificial-intelligence-and-machine-learning/3/#383de2aa2bfc

Deep Learning is used by Google in its voice and image recognition

algorithms, by Netflix and Amazon to decide what you want to watch
or buy next, and by researchers at MIT to predict the future. The ever-
growing industry which has established itself to sell these tools is
always keen to talk about how revolutionary this all is. But what
exactly is it? And is it just another fad being used to push
“old fashioned” AI on us, under a sexy new label?
* What Is The Difference Between Deep Learning, Machine Learning and AI?, Forbes, Dec 8, 2016
https://www.forbes.com/sites/bernardmarr/2016/12/08/what-is-the-difference-between-deep-learning-machine-learning-and-ai/#36fb681d26cf
11
 02 Machine Learning(ML)

“Some Studies in Machine Learning Using the Game of Checkers“

A. L. Samuel, IBM Journal of Research and Development
( Volume: 3, Issue: 3, July 1959 )
* http://ieeexplore.ieee.org/document/5392560/,
https://en.wikipedia.org/wiki/Machine_learning

[Definition 1] Arthur Samuel(1959)

Field of study that gives computers the ability
to learn without being explicitly programmed.

[Definition 2] Tom Mitchell(1998)

A computer program is said to learn from experience
E with respect to some task T and some performance
measure P, if its performance on T, as measured by P,
improves with experience E.
* Andrew Ng, Coursera Machine Learning lecture(https://www.coursera.org/learn/machine-learning/), Lecture 1
* Robot image : https://www.aitimejournal.com/@premlatha.kr/what-is-ai-in-a-simple-way 12
ML vs Traditional Program

Traditional Program
Program Output
Input
(Rule)

Machine Learning
Input (Training ML Trained Model
Answer
Data)
Model (Rule)

(New) Trained Output

Input Model(Rule)
* Important factors for Machine Learning:
①Input ②Output(Answer) ③Model
13
 ML vs Traditional Program
➢ Traditional Program
Explicitly program “Output X if B is true with input A”
➢ Machine Learning
Train an algorithm to find condition B
with (input A, answer X) pair

Traditional Program Machine Learning

Calculate following expressions. What’s the number for □ and △?

▪ 3 x 2= ✓ input A : numbers ▪ 3 x □ + 2 x △ = 1 Condition B

▪ 4 x -3 = ✓ condition B : x ▪ 1 x □ + 4 x △ = -3 □= 1
▪ 5 x 8= (multiplication) ▪ 5 x □ + 5 x △ = 0 △ = -1
▪ 7 x -6 = ✓ output X : answer ▪ 8 x □ + 3 x △ = 5

Input A Answer X
* https://www.slideshare.net/JinwonLee9/ss-70446412
* Andrew Ng , Coursera Machine Learning 강의(https://www.coursera.org/learn/machine-learning/), Lecture 1
* “구글에서 배우는 딥러닝” 28쪽, 닛케이 빅 데이터 著, 서재원 옮김, 영진닷컴(Book: Deep learning from Google] 14
 Estimating
ML vs Traditional Program House price

Traditional Program Machine Learning

price = 0 Machine learning is
price_per_sqft = 200 calculating the weight of
if neighborhood == 'Good' each factor
price_per_sqft = 400
elseif neighborhood == 'Bad'
price_per_sqrt = 100 price =

price = price_per_sqft * sqft neighborhood * 2.3

if num_of_bedrooms == 1 + sqft * 1,231

price = price - 20,000
else + num_of_bedrooms * 0.8
price = price +
(num_of_bedrooms * 1,000) + 201

15
* https://medium.com/@ageitgey/machine-learning-is-fun-80ea3ec3c471 (Modified)
 Areas of Machine Learning
➢ Supervised Learning
➢ Training with (input, label) pair
➢ Classification, Regression
➢ (Model) SVM, Decision Tree, kNN, Regression
➢ Unsupervised Learning
➢ Training only with input (w/o label)
➢ Clustering, Compression
➢ (Model) K-means clustering
➢ Reinforcement Learning
➢ Reward instead label
➢ Action Selection, Policy Learning
➢ (Model) MDP(Markov Decision Process)
* https://www.slideshare.net/JinwonLee9/ss-70446412
* Andrew Ng, Coursera Machine Learning(https://www.coursera.org/learn/machine-learning/), Lecture 1, 2
* 인공지능, 머신러닝, 딥러닝 입문, 김의중, 위키북스(Book: Introduction to AI, ML, DL) 16
 Supervised Learning
Example of ML – Linear regression

Estimating food truck profit

According to population of city

* Andrew Ng, Coursera Machine Learning(https://www.coursera.org/learn/machine-learning/), Programming Exercise 1, Lecture 2 17

 Supervised Learning
Example of ML – Linear regression

(Core algorithm)
for iter = 1:num_iters
predictions = X * theta;
theta_0 = theta(1) - alpha / m *
sum(predictions - y);
theta_1 = theta(2) - alpha / m *
(X(:,2)' * (predictions - y));
theta = [theta_0; theta_1];
end
* Andrew Ng, Coursera Machine Learning lecture(https://www.coursera.org/learn/machine-learning/), Programming Exercise 1, Lecture 2 18
 Supervised Learning
Example of ML – Linear regression

https://www.youtube.com/watch?v=4PHI11lX11I
Linear Regression using Gradient Descent in Python - Machine Learning Basics(00:09)
19
 Unsupervised Learning
Example of ML – Clustering

20
* https://www.youtube.com/watch?v=bHi6DacI3os, Machine Learning: What is k-Means clustering?(01:46)
 With
Human teaching
Example of ML – Reinforcement Learning

Table Tennis

https://www.youtube.com/watch?v=SH3bADiB7uQ
Towards Learning Robot Table Tennis(02:37), 2012. 5. 24.
21
 Without
Human teaching
Example of ML – Reinforcement Learning

Block break game

https://www.youtube.com/watch?v=V1eYniJ0Rnk
Google DeepMind's Deep Q-learning
playing Atari Breakout(01:42), 2015.3.7.
22
 03 Artificial Neural Network(ANN)
ANN is based on a collection of
connected units or nodes called
artificial neurons, which loosely
model the neurons in a biological
brain. (Wikipedia)

Artificial neuron:
neuron in the brain Perceptron(Rosenblatt, 1958)

* http://solarisailab.com/archives/1206
* https://en.wikipedia.org/wiki/Biological_neuron_model
*https://www.yumpu.com/en/document/view/33756198/axon-dendrite-cell-body-neuron-nerve-cell-anatomy-synapse 23
 ANN, Multi-Layer Perceptron(MLP)

Artificial Neural Network Backpropagation

Multi-Layer Perceptron(MLP) (1986, Geoffrey Hinton et al.)
(1986, Geoffrey Hinton et al.)

* http://solarisailab.com/archives/1206 24
 Example of ANN – Autonomous Vehicle

* 2017년 특허청 심사관 신기술 교육(제4차 산업혁명과 인공지능) 교재[Book: New technology textbook for Patent examiner] 25
 Example of ANN – Autonomous Vehicle
➢ ALVINN: Autonomous Land Vehicle In a Neural Network
▪ First developed in 1986 by Carnegie Mellon University’s NavLab
▪ Architecture: Single hidden layer back-propagation network,
two retinas and a single intensity feedback unit

* https://papers.nips.cc/paper/95-alvinn-an-autonomous-land-vehicle-in-a-neural-network.pdf
* https://www.youtube.com/watch?v=5-acCtyKf7E 26
 04 Deep Learning(DL)
Deep learning is part of a broader
family of machine learning methods
based on artificial neural networks
with representation learning(feature
learning). (Wikipedia)

* http://solarisailab.com/archives/1206

* https://www.analyticsvidhya.com/blog/2017/04/comparison-between-deep-learning-machine-learning/

27
 Event 1 of Deep Learning

* Yoshua Bengio “Scaling up Deep Learning”(2014), http://www.iro.umontreal.ca/~bengioy/talks/KDD2014-tutorial.pdf 28

 Event 2 of Deep Learning

SuperVision team
(AlexNet) deep convolutional
Alex Krizhevsky, Ilya Sutskever, neural network
Geoffrey Hinton
University of Toronto

* Fei-Fei Li : Lecture 1 | Introduction to Convolutional Neural Networks for Visual Recognition,

https://www.youtube.com/watch?v=vT1JzLTH4G4
* http://www.image-net.org/challenges/LSVRC/2012/results.html * http://www.dataversity.net/brief-history-deep-learning/ 29
 Event 3 of Deep Learning

< Google, 2012, Andrew Ng el al. >

✓ Image recognition by unsupervised learning
* Trained the model with 10 million of youtube capture image(200x200) for 3 days
* Used 16,000 CPU core(server 1000), 1 billion parameter
* Accuracy : Human face 81.7%, Human body 76.7%, Cat 74.8%
* https://www.analyticsvidhya.com/blog/2017/04/comparison-between-deep-learning-machine-learning/
* http://www.hani.co.kr/arti/science/science_general/777909.html
* https://blog.google/topics/machine-learning/using-large-scale-brain-simulations-for/
* https://static.googleusercontent.com/media/research.google.com/ko//archive/unsupervised_icml2012.pdf 30
 Concept of Deep Learning(DL)
➢ DL is what overcame the limitations of ANN
➢ [Problem] Hard to train → [Solution] Pretrain
➢ [Problem] Difficult to handle new data

→ [Solution] Dropout training data intentionally

* Both are solved by Geoffrey Hinton(2006, 2012)
➢ DL is just a new brand different from ANN
People didn’t like ANN and thought it would not be
➢

possible to use ANN again. (Reason of 2nd AI winter)

Success Algorithm Big Data Good H/W
Overcame Rich Enhanced GPU,
factor = limitations of + training + Computation
of DL ANN data power up
* http://economyplus.chosun.com/special/special_view_past.php?boardName=C24&t_num=11363&img_ho=190
* “인간 vs 기계” , 김대식, 동아시아[Book: Human vs Machine]
* https://www.youtube.com/watch?v=MpUNjGeon70
* “On the Origin of Deep Learning” 3page, Haohan Wang, Bhiksha Raj, https://arxiv.org/pdf/1702.07800.pdf 31
 Deep vs Shallow
At which problem depth does Shallow Learning end, and Deep
Learning begin? Discussions with DL experts have not yet yielded
a conclusive response to this question. […], let me just define for
the purposes of this overview:
problems of depth > 10 require Very Deep Learning.
* https://machinelearningmastery.com/what-is-deep-learning/

* https://www.slideshare.net/JinwonLee9/ss-70446412
32
 DL, Self feature engineering
Deep learning allows computational models to learn
representations of data with multiple levels of
abstraction.
Deep-learning methods are representation-learning
methods with multiple levels of representation,
obtained by composing simple but non-linear modules
that each transform the representation at one level
(starting with the raw input) into a representation at
a higher, slightly more abstract level. […] The key
aspect of deep learning is that these layers of
features are not designed by human engineers: they
are learned from data using a general-purpose
learning procedure.
* https://machinelearningmastery.com/what-is-deep-learning/
* Deep learning(Yann LeCun, Yoshua Bengio, Geoffrey Hinton), Nature 521, 436-444(28 May 2015), 10.1038/nature14539 33
 Feature extraction

* “Introduction to Pattern Analysis”, http://www.nada.kth.se/~stefanc/DATORSEENDE_AK/pr_l1.pdf

34
 Difficulty of feature extraction
What’s the rule to recognize below cats? Can we find it?

* Fei-Fei Li : Lecture 2 | Image Classification, https://www.youtube.com/watch?v=OoUX-nOEjG0 35

 Difficulty of feature extraction

* Fei-Fei Li : Lecture 2 | Image Classification Lecture, https://www.youtube.com/watch?v=OoUX-nOEjG0 36

Deep Learning vs Machine Learning

* Deep Reinforcement Learning, Lecture 1, 14p, CS 294-112 at UC Berkeley,

http://rail.eecs.berkeley.edu/deeprlcourse/static/slides/lec-1.pdf 37
 Convolutional
Representative model of DL - CNN Neural
Network

Convolution

Pooling

* An Intuitive Explanation of Convolutional Neural Networks, https://ujjwalkarn.me/2016/08/11/intuitive-explanation-convnets/ 38

 CNN - 이미지 인식 최강

머신러닝 딥러닝 CNN

(HOG+SVM) * 출처 : Fei-Fei Li : Lecture 1 | Introduction to Convolutional Neural Networks for Visual Recognition 강의자료 39분, 39
https://www.youtube.com/watch?v=vT1JzLTH4G4
 Recurrent
Representative model of DL - RNN Neural
Network

Machine Translation: use 2 RNNs

* http://www.wildml.com/2015/09/recurrent-neural-networks-tutorial-part-1-introduction-to-rnns/
40
* https://medium.com/@ageitgey/machine-learning-is-fun-part-5-language-translation-with-deep-learning-and-the-magic-of-sequences-2ace0acca0aa
 CNN + RNN for image description

* http://www.wildml.com/2015/09/recurrent-neural-networks-tutorial-part-1-introduction-to-rnns/ 41
* https://medium.com/@ageitgey/machine-learning-is-fun-part-5-language-translation-with-deep-learning-and-the-magic-of-sequences-2ace0acca0aa
 DL needs Training/Retraining

“When trained on a new task, standard neural networks forget

most of the information related to previously learned tasks, a
phenomenon referred to as ‘catastrophic forgetting’.”
* Three scenarios for continual learning, https://arxiv.org/pdf/1904.07734.pdf

Forget
everything?

42
 DL needs Training/Retraining
Classification using DL

Bike

What if,
1. To add a new class?
(ex) Want to classify bike
→ Need to train whole
model again

2. Quality decline?
(ex) The model cannot
classify new car models
→ Train new models only
or whole model
* Andrew Ng, Coursera Machine Learning lec(https://www.coursera.org/learn/machine-learning/), Lecture 8
* Bike image : https://kr.clipart.me/transportation/bike-bicycle-clip-art-44237 43
 Retraining vs Continuous Learning

Retraining Continuous Learning

Because data distributions can be (Life-long learning)
expected to drift over time, deploying
1. Catastrophic forgetting:
a model is not a one-time exercise but
When trained on a new task,
rather a continuous process. It is a
standard neural networks forget
good practice to continuously monitor
most of the information related
the incoming data and retrain your
to previously learned tasks.
model on newer data. If monitoring has
a high overhead, then a simpler 2. Continuous Learning:
strategy is to train the model enable a model to learn
periodically. continuously without forgetting.
* Model Drift refers to a model’s predictive
(The model leverage its past
performance degrading over time due to a experience.)
change in the environment. (Measuring the
accuracy of a deployed model on live data is * https://arxiv.org/abs/1904.07734
a notoriously difficult problem.) * https://ruder.io/transfer-learning/

* https://docs.aws.amazon.com/machine-learning/latest/dg/retraining-models-on-new-data.html
* https://mlinproduction.com/model-retraining/ 44
 Strength of Deep Learning

* https://www.quora.com/What-is-the-difference-between-Deep-Learning-Machine-learning-and-Artificial-Intelligence-Is-Deep-learning-related-to-data-science
45
Godfathers of Deep Learning

Yann LeCun, Geoffrey Hinton, Yoshua Bengio, Andrew Ng

Three won Turing Award 2018 this year for their work on artificial neural networks.
(Nobel Prize of computing)
* https://www.kdnuggets.com/2015/03/talking-machine-deep-learning-gurus-p1.html
* https://www.youtube.com/watch?v=HzilDIhWhrE 46
05 AI-related IP issues
IP : Intellectual Property

“AI tools allow us to achieve better quality

and reduce administration costs.”
- WIPO Director General Francis Gurry, Sep. 2018

Can AI be an inventor of a patent?

Can AI be an author of an artistic work?

https://www.wipo.int/wipo_magazine/en/2018/05/article_0001.html
Robot image: https://publicdomainvectors.org/ko/tag/%EB%A1%9C%EB%B4%87/2 47
Automation issue

AI and IP automation

 Why do IP offices need AI in the administration?

 Volume is a principal driver: The volume of prior art is rapidly
exceeding the processing capacity of available human resources

 Quality and cost are also important drivers

 Challenges for IP offices

 Building AI capacity is a major challenge for all IPOs

 Professionals with required knowledge is limited

 International collaboration and cooperation is needed

https://www.wipo.int/wipo_magazine/en/2018/05/article_0001.html
48
Automation issue

Where to apply AI for IP offices?

 Customer Service  Classification

 Chabot, Voice-based  Patent, Trademark,
customer service system Design

 Patent Prior art Search  Image Search

 Automatic search  Trademark, Design
system  Patent drawing

 Machine Translation  Other areas

 Neural Machine  STT, OCR
Translation (NMT)  Formality check
49
Patent policy issue

Patent application invented by AI

Device for the

Autonomous
Bootstrapping of
Unified
Sentience.

* https://www.bbc.com/news/technology-49191645 50
 Patent policy issue

Patent application invented by AI

[1. Food Container]

[2. Devices and Methods for

Attracting Enhanced Attention]

* applicant’s homepage: http://artificialinventor.com/patent-applications/

* Reasons for refusal: https://www.epo.org/news-events/news/2020/20200128.html
* Food container: https://register.epo.org/application?number=EP18275163#_blank
* Attracting device: https://register.epo.org/application?number=EP18275174#_blank 51
Patent policy issue Ryan Abbott and his team

What they are arguing about DABUS

✓A machine architecture called “DABUS” conceived of the instant inventions
✓Inventorship should not be restricted to natural persons.
A machine that would meet inventorship criteria if it were a natural person
should also qualify as an inventor.
* Inventorship is restricted to individuals under, e.g., 35 U.S.C. §100(f) (1952)
✓An “autonomous machine invention” should be assigned
to the owner of the machine
* Machines should not own patents. They do not have legal personality
or independent rights, and cannot own property.
✓The machine’s owner is prevented from listing himself as the inventor
for the instant application.
* Inaccurately listing himself as an inventor could subject him to criminal sanctions.
✓The first step of determination of ownership has to be the identification of
the inventor(true deviser of the invention), as without knowing the inventor
one cannot determine who owns the invention.
✓Patent laws already provide for the rights in an invention to vest with
someone else in place of the inventor.(successor in title, etc)
* http://artificialinventor.com/patent-applications/
* https://www.surrey.ac.uk/people/ryan-abbott 52
Patent policy issue

AI patent application status

http://artificialinventor.com/

✓ Refused: EPO, USPTO, UKIPO

✓ Gazette Published: WIPO

53
Patent policy issue

IP5 expert round table on AI invention

 Inventorship/ownership (Programmer, User, etc)

 All IP5 require that the inventor is a human being.
 Difficult to determine whether by human or machine
 Patent eligibility
 Can be handled as CII(Computer-Implemented Inventions)
 Sufficiency of disclosure
 Disclose how the model was trained and provides the
data used for training
 Inventive step
 Currently, no need to be modified for AI inventions

* https://www.fiveipoffices.org/material/ai_roundtable_2018_report/ai_roundtable_2018_report/ai_roundtable_2018_report 54
Copyright policy issue

AI Arts

* https://deepdreamgenerator.com/ * https://www.jwt.com/en/work/thenextrembrandt * https://aiartists.org/ 55

Copyright policy issue

Same issues with copyright

 Authorship/ownership
 Only works created by human can be protected

 Designating a human or Public domain(Free of copyright)

 UK: In the case of an artistic work which is computer-generated, the

author shall be the person by whom the arrangements necessary for the
creation of the work are undertaken.

 level of Originality ✓ It can be solved on a case-by-case basis.

✓ MS Word: MS does not own every piece
of work produced using it.
 Protecting dataset ✓ AI: Sometimes the user’s contribution
to the creative process may simply be
to press a button.
* https://mediasvr.egentouch.com/egentouch.media/fileHandler.do?action=fileViewDown&SCHOOL_ID=1010023&PMEDIA_ID=103189
* https://www.wipo.int/wipo_magazine/en/2017/05/article_0003.html 56
 Copyright policy issue

Copyright case with animal

Monkey: Naruto Wikipedia posted the selfies on it’s free-to-

Photographer: use website. Slater asked that it be taken
David John Slater down. Wikipedia argued the photo is
uncopyrightable because an animal took it.

➢ That court upheld a lower court's previous ruling, which said,

basically, that animals can't file copyright infringement suits.
➢ In 2011, Naruto, then 7 years old, snapped several photos of
himself with a camera belonging to photographer David John Slater,
who was on assignment in Indonesia.
➢ Slater included Naruto's photos in a book he published.
* https://edition.cnn.com/2018/04/24/us/monkey-selfie-peta-appeal/index.html
57
 How to solve? Back to basics
What’s the purpose of the law?
How to achieve it?
A starting point for developing IP policy, and indeed innovation
policy, is to identify desired outcomes. What is it that we
want to achieve? This is the fundamental question. If society
considers innovation will be encouraged by giving a
right to a machine, then such a proposition may gain
traction. - WIPO Director General Francis Gurry, Oct. 2019

The purpose of (Korean) patent and copyright law

➢ PATENT: Encourage, protect and utilize inventions to promote the
development of technology and industry.
➢ COPYRIGHT: Protect the rights of authors and promote fair use of works in
order to contribute to the improvement and development of culture and
related industries.
* https://www.wipo.int/wipo_magazine/en/2019/05/article_0001.html
* http://www.law.go.kr/lsInfoP.do?lsiSeq=192474&urlMode=engLsInfoR&viewCls=engLsInfoR#0000 58
 International meeting

1. Opening panel 2. Patent panel

3. Governance and Development panel 4. Copyright panel
5. Data panel 6. IPO Administration panel
* https://www.wipo.int/about-ip/en/artificial_intelligence/policy.html

IP5 NET/AI Task force

Representatives of the IP5 Offices and WIPO gathered
at the inaugural meeting of their joint task force on new
emerging technologies (NET) and artificial intelligence
(AI) in Berlin on 15 and 16 January 2020.
* https://www.fiveipoffices.org/news/20200117

59
 The more you know about AI,
the better you can do with it.

Thank you.