Law and technology
Part I. Technology, economy & society
Chapter 1. An introduction to information technology for lawyers
Section 1. introduction
Technology is neither good nor bad, nor is it neutral
- Lees: Melvin Kranzberg (1986) Technology and History: “Kranzberg’s Laws”.
Technology and Culture, 27(3), 544-560. doi:10.2307/3105385
§2. Technology, society, economy
All three influence each other
- Societies needs are the basis of new technological changes, which impacts the
economy. And you can say the exact same thing the other way around
§3. Summary
How does the technology work?
- How is the hardware (physical parts) / software (instructions / code) designed?
- Which models are used?
- How are models trained?
- Which data is used?
Who has designed the technology?
How well does the technology work?
How secure is the technology?
Section 2. how does technology work
§1. Hardware vs. software
Computer systems are built of two components
Hardware
- All physical parts
- Harddrives, memory keyboard,…
Software
- All instructions and codes that make hardware work
§2. Which models are used
The internet exists thanks to big data architecture
- Uses: camera sensors, internet traffic, …
- Thanks to algorithms
o Is an sequence of instructions that describes how to realise a goal
o Is a code that gathers information and derives it in a specific way (used for
the needed goals)
These techniques are characterized by 5V’s
- Volume: vast amount of data that it processes
- Velocity: speed which it generates data
o It has to adapt to each situation and data change
, - Variety: two main types
o Structured: data stored in a table or relational database, easy to gather
o Unstructured: more difficult to gather and process
Images, videos, sound, social media texts, radiographical images, …
- Veracity: quality of sources and informations can vary
- Value: common goal is to value these data
o Give meaning to it
o Scientific domains, business to calculate better company traject, economy
and sustainability
o Law enforcement: analyzing of criminal records and following everything
§3. Artificial intelligence
Basis blocks to be AI
- Interact with environment
o By analyzing the world around itself (collecting and processing information)
using camera’s to gather and process info
- Reason and plan in the real world
o Ex. Route planner has to make a good decision for the best route
- Ability to learn and adapt
o Learn from mistakes
An intelligent agent is a system that can fulfill a function in its environment and learn and
adapt in that environment
§4. Which models are used and trained
Machine learning: ML models
- Large amount of data and examples that make a computer able to learn certain
things itself, no need to specific programming
- Predictive ML-model
o Comes to a specific outcome by analysing the data (input)
o It needs to be trained with examples that are labelled in order for it to
distinguish each thing from another
Diagnosing a sickness
o Constructing the model with examples = training it
o Danger: overfitting
ð not being able to generalize beyond it
- Unsupervised ML-model
o Finding structures in data
Spotify music taste
- Reinforcement ML-model
o Aim to learn a control policy
o They have a goal and learn the optimal way to reach it by learning by itself
and positive feedback
Neural networks and deep learning
- Artificial neural network (ANN): mathematical abstraction to link things (aka
neurons in brain)
- ð Multi-layer perceptron (MLP): techniques needs to be combined to make a better
AI
- Renewed interest
o 1. Better progressing speed of data
o 2. Better image analysis in order to process larger and more complex types
- Disadvantages
, o Need lots of data to train
o Need lots of computational power to train it
o Built on the data and not possible to know what the data is (for users)
It’s difficult for humans tot trust these models
o Vulnerable to adverbial attacks
Natural language processing
- The Turing test: system is intelligent if a human cannot distinguish it from another
human
- Needs
o Answering correctly, remembering historical conversations, knowledge
about the real world, generate sensible answers
o Good language structure, writer and spoken
o Learning if their way of formulating it is positively evaluated by the human
- Now it’s very realistic
o Only thing missing: understanding the semantic of the output and not only
giving an answer based on statistics
Section 3. cryptocurrencies, blockchain, cryptography
Enhancing trust in these systems
- Used for online banking and accessing panels
- Public-private key
Blockchain and cryptocurrencies
- Reaction to power of central banks
- How to trust
o No central authority: trust comes from mathematical properties
o Identity management: personal keys and privacy from pseudonymity
o Same principles as classic monetary systems
Enough money, no double transactions, …
o No money accumulation, but chain transaction results
- Spendings controlled by blockchain
o No double spending
o Mining
Section 4. impact on society
Changes the way of communication, business, entertainment, information consumption,
…
Great promise, with caution
- Attacks or hacking
- The panacea view: it is not neutral, so beware
o A lot of discrimination in the data, so the output is also discriminating
Need to cooperate with AI, rather than letting it replace us
Chapter 2. The rise of platforms in the digital economy
, Section 1. introduction
2011 – “Why software is eating the world” – Marc Andreessen
- Predicted that large parts of the economy would be software-enterprises
ð came true: Amazon, Google, Meta, …
Creative destruction (Schumpeter)
- Came true: new structures replace older less innovative ones
Exponential growth of digital economy
Centrality of digital platforms is big: 7 largest companies are digital markets
- Influence our social behaviour, economy, entertainment, …
- Covid-19 helped this
o More technology advances with governmental help
Section 2. platforms as a business model
Term
- Originally: a set of technical specifications or (building blocks’ upon which third
parties could develop products, services or technologies
o This structure helped others to develop business models more easily
Booksellers with ISBN, barcodes, …
Platforms became actual business models itself
- Traditional models: enterprises focus on transformation of raw materials into a
product
- Digital platform: provide services
Digital platform
- Central idea: bringing two or more groups together to connect, interact,…
o At least two
- Success is efficiency matchmaking
o Netflix algorithm
- They facilitate interactions and transactions
Intermediate position between supply and demand
- The more market groups the more complicated the business model
- 1. Two-sided: bringing two sides of groups together
o Visa: payment between customer and seller
o Uber
- 2. Multi-sided
o Facebook: app developers, advertisers, users, …
- 3. Platform ecosystems: interconnects multiple platforms
o Google, Alphabet, Amazon
Section 3. platform characteristics
§1. Use of digital technology
The use of special software technology makes the connection users easier and attracts
them
- Attracting by matchmaking
Purely digital
Part I. Technology, economy & society
Chapter 1. An introduction to information technology for lawyers
Section 1. introduction
Technology is neither good nor bad, nor is it neutral
- Lees: Melvin Kranzberg (1986) Technology and History: “Kranzberg’s Laws”.
Technology and Culture, 27(3), 544-560. doi:10.2307/3105385
§2. Technology, society, economy
All three influence each other
- Societies needs are the basis of new technological changes, which impacts the
economy. And you can say the exact same thing the other way around
§3. Summary
How does the technology work?
- How is the hardware (physical parts) / software (instructions / code) designed?
- Which models are used?
- How are models trained?
- Which data is used?
Who has designed the technology?
How well does the technology work?
How secure is the technology?
Section 2. how does technology work
§1. Hardware vs. software
Computer systems are built of two components
Hardware
- All physical parts
- Harddrives, memory keyboard,…
Software
- All instructions and codes that make hardware work
§2. Which models are used
The internet exists thanks to big data architecture
- Uses: camera sensors, internet traffic, …
- Thanks to algorithms
o Is an sequence of instructions that describes how to realise a goal
o Is a code that gathers information and derives it in a specific way (used for
the needed goals)
These techniques are characterized by 5V’s
- Volume: vast amount of data that it processes
- Velocity: speed which it generates data
o It has to adapt to each situation and data change
, - Variety: two main types
o Structured: data stored in a table or relational database, easy to gather
o Unstructured: more difficult to gather and process
Images, videos, sound, social media texts, radiographical images, …
- Veracity: quality of sources and informations can vary
- Value: common goal is to value these data
o Give meaning to it
o Scientific domains, business to calculate better company traject, economy
and sustainability
o Law enforcement: analyzing of criminal records and following everything
§3. Artificial intelligence
Basis blocks to be AI
- Interact with environment
o By analyzing the world around itself (collecting and processing information)
using camera’s to gather and process info
- Reason and plan in the real world
o Ex. Route planner has to make a good decision for the best route
- Ability to learn and adapt
o Learn from mistakes
An intelligent agent is a system that can fulfill a function in its environment and learn and
adapt in that environment
§4. Which models are used and trained
Machine learning: ML models
- Large amount of data and examples that make a computer able to learn certain
things itself, no need to specific programming
- Predictive ML-model
o Comes to a specific outcome by analysing the data (input)
o It needs to be trained with examples that are labelled in order for it to
distinguish each thing from another
Diagnosing a sickness
o Constructing the model with examples = training it
o Danger: overfitting
ð not being able to generalize beyond it
- Unsupervised ML-model
o Finding structures in data
Spotify music taste
- Reinforcement ML-model
o Aim to learn a control policy
o They have a goal and learn the optimal way to reach it by learning by itself
and positive feedback
Neural networks and deep learning
- Artificial neural network (ANN): mathematical abstraction to link things (aka
neurons in brain)
- ð Multi-layer perceptron (MLP): techniques needs to be combined to make a better
AI
- Renewed interest
o 1. Better progressing speed of data
o 2. Better image analysis in order to process larger and more complex types
- Disadvantages
, o Need lots of data to train
o Need lots of computational power to train it
o Built on the data and not possible to know what the data is (for users)
It’s difficult for humans tot trust these models
o Vulnerable to adverbial attacks
Natural language processing
- The Turing test: system is intelligent if a human cannot distinguish it from another
human
- Needs
o Answering correctly, remembering historical conversations, knowledge
about the real world, generate sensible answers
o Good language structure, writer and spoken
o Learning if their way of formulating it is positively evaluated by the human
- Now it’s very realistic
o Only thing missing: understanding the semantic of the output and not only
giving an answer based on statistics
Section 3. cryptocurrencies, blockchain, cryptography
Enhancing trust in these systems
- Used for online banking and accessing panels
- Public-private key
Blockchain and cryptocurrencies
- Reaction to power of central banks
- How to trust
o No central authority: trust comes from mathematical properties
o Identity management: personal keys and privacy from pseudonymity
o Same principles as classic monetary systems
Enough money, no double transactions, …
o No money accumulation, but chain transaction results
- Spendings controlled by blockchain
o No double spending
o Mining
Section 4. impact on society
Changes the way of communication, business, entertainment, information consumption,
…
Great promise, with caution
- Attacks or hacking
- The panacea view: it is not neutral, so beware
o A lot of discrimination in the data, so the output is also discriminating
Need to cooperate with AI, rather than letting it replace us
Chapter 2. The rise of platforms in the digital economy
, Section 1. introduction
2011 – “Why software is eating the world” – Marc Andreessen
- Predicted that large parts of the economy would be software-enterprises
ð came true: Amazon, Google, Meta, …
Creative destruction (Schumpeter)
- Came true: new structures replace older less innovative ones
Exponential growth of digital economy
Centrality of digital platforms is big: 7 largest companies are digital markets
- Influence our social behaviour, economy, entertainment, …
- Covid-19 helped this
o More technology advances with governmental help
Section 2. platforms as a business model
Term
- Originally: a set of technical specifications or (building blocks’ upon which third
parties could develop products, services or technologies
o This structure helped others to develop business models more easily
Booksellers with ISBN, barcodes, …
Platforms became actual business models itself
- Traditional models: enterprises focus on transformation of raw materials into a
product
- Digital platform: provide services
Digital platform
- Central idea: bringing two or more groups together to connect, interact,…
o At least two
- Success is efficiency matchmaking
o Netflix algorithm
- They facilitate interactions and transactions
Intermediate position between supply and demand
- The more market groups the more complicated the business model
- 1. Two-sided: bringing two sides of groups together
o Visa: payment between customer and seller
o Uber
- 2. Multi-sided
o Facebook: app developers, advertisers, users, …
- 3. Platform ecosystems: interconnects multiple platforms
o Google, Alphabet, Amazon
Section 3. platform characteristics
§1. Use of digital technology
The use of special software technology makes the connection users easier and attracts
them
- Attracting by matchmaking
Purely digital
