Courses
Managing the Risk of Intelligent Algorithms
Advanced seminar (SS 18) - Organisation
News
Summary
Artificial intelligence based on machine-learning leads to a new type of algorithm: The intelligent algorithm. Intelligent algorithms have a tremendous positive potential in many application fields, some already alive today. On the other hand, intelligent algorithms introduce new risks, with potentially very harmful outcomes. Managing, i.e. understanding, assessing, and mitigating those risks becomes an important part when developing, implementing and operating intelligent algorithms.
Context
With the successful rise of machine-learning, a new type of algorithm became possible: The intelligent algorithm. Intelligent algorithms have the capability to learn from their environment, from their actions, from big data, and from other sources.
Intelligent algorithms:
The behaviour of intelligent algorithms is already more successful in a number of areas than their human counterpart. Thus, the rise of intelligent algorithms is assured. Civilization can expect higher levels of quality of life, safety, and efficiency through the use of intelligent algorithms.
However, the capabilities of intelligent algorithms also introduce new, sometimes significant, risk. Because they often autonomously control cyber-physical systems, physical damage to life or property is possible. Due to their decision capabilities based on big data, they may influence decisions on human health, employment chances, investment opportunities, traffic control, … or other areas important for humans or for society.
Intelligent algorithms generate:
Our task as engineers is to find and contribute to the acceptable balance between benefits and risks. This is especially true in the field of intelligent algorithms and leaves us with a high responsibility.
Intelligent algorithms:
- Have a high degree of autonomy, due to their capability of self-learning;
- Take decisions based on the state of the environment and their behaviour, and improve by self-learning;
- Often control cyber-physical systems, i.e. interact with the physical world.
The behaviour of intelligent algorithms is already more successful in a number of areas than their human counterpart. Thus, the rise of intelligent algorithms is assured. Civilization can expect higher levels of quality of life, safety, and efficiency through the use of intelligent algorithms.
However, the capabilities of intelligent algorithms also introduce new, sometimes significant, risk. Because they often autonomously control cyber-physical systems, physical damage to life or property is possible. Due to their decision capabilities based on big data, they may influence decisions on human health, employment chances, investment opportunities, traffic control, … or other areas important for humans or for society.
Intelligent algorithms generate:
- Technical risks (Malfunction, loss of service, incorrect decisions, …);
- Legal risks (Assignment of responsibility, liability, legal conformance, …);
- Ethical risks (Moral questions, boundary between artificial and human intelligence, supervision, …).
Our task as engineers is to find and contribute to the acceptable balance between benefits and risks. This is especially true in the field of intelligent algorithms and leaves us with a high responsibility.
Seminar Work
This seminar will work on the central theme: How can we manage i.e. recognize, assess, and mitigate the risks of intelligent algorithms?
Each participant choses one of the 3 fields:
F1: Which are possible technical risks when developing and operating intelligent algorithms? How can we manage these (or a subset)?
F2: Which are possible legal risks when developing and operating intelligent algorithms? How can we manage these (or a subset)?
F3: Which are possible ethical risks when developing and operating intelligent algorithms? How can we manage these (or a subset)?
The Hauptseminar has 3 seminar days (see separate work program, dates below):
Each participant choses one of the 3 fields:
F1: Which are possible technical risks when developing and operating intelligent algorithms? How can we manage these (or a subset)?
F2: Which are possible legal risks when developing and operating intelligent algorithms? How can we manage these (or a subset)?
F3: Which are possible ethical risks when developing and operating intelligent algorithms? How can we manage these (or a subset)?
The Hauptseminar has 3 seminar days (see separate work program, dates below):
- An introduction day: Intelligent Algorithms will be introduced in a lecture by Professor Dr. Frank J. Furrer, and the parts of the Hauptseminar (Paper, presentation) will be defined,
- Individual, guided research in the selected area and authoring of a scientific paper. Feedback from peer reviewers,
- A first seminar day: The participants will present their results and receive feedback from the audience,
- Improvement of the paper and the presentation, based on the peer feedback,
- A second seminar day: The participants will present their improved results and receive feedback from the audience,
- Delivery of the final paper.
Learning Outcome
The participants will learn: (a) to do focused research in a specific area (Intelligent Algorithms”), (b) to author a scientific paper, (c) to improve their LATex expertise, (d) to experience the peer-review process and (e) to hold convincing presentations, and (f) to benefit from a considerable broading of their perspective in the field of technology, software, and applications.
Seminar language is English. Three seminar days will be held and 3 ECTS credits are awarded for the successful participation.
Audience is limited to 7 participants. Please register in advance.
Seminar language is English. Three seminar days will be held and 3 ECTS credits are awarded for the successful participation.
Audience is limited to 7 participants. Please register in advance.
Mandatory Reading
[1] Managing algorithmic risks Safeguarding the use of complex algorithms and machine learning. 2017 Deloitte Development LLC., New York, USA. Downloadable from: Link [last accessed: 7.2.2018]
[2] Demystifying artificial intelligence What business leaders need to know about cognitive technologies. 2014 Deloitte Development LLC., New York, USA. Downloadable from: Link [last accessed: 7.2.2018]
[3] James Barrat: Our Final Invention Artificial Intelligence and the End of the Human Era. Griffin Publishing, 2015. ISBN 978-1-2500-5878-2
[2] Demystifying artificial intelligence What business leaders need to know about cognitive technologies. 2014 Deloitte Development LLC., New York, USA. Downloadable from: Link [last accessed: 7.2.2018]
[3] James Barrat: Our Final Invention Artificial Intelligence and the End of the Human Era. Griffin Publishing, 2015. ISBN 978-1-2500-5878-2
Seminar Schedule
Kick-Off Meeting (Introduction): Wednesday, April 25, 2018 / 11:10 12:40 in APB/INF 2101
Seminar Day 1: Wednesday, June 13, 2018 / 09:20 10:50 & 11:10 12:40 in APB/INF 2101
Seminar Day 2: Wednesday, July 11, 20187 / 09:20 10:50 & 11:10 12:40 in APB/INF 2101
Seminar Day 1: Wednesday, June 13, 2018 / 09:20 10:50 & 11:10 12:40 in APB/INF 2101
Seminar Day 2: Wednesday, July 11, 20187 / 09:20 10:50 & 11:10 12:40 in APB/INF 2101
Allowances
The course can be used for the modules as specified by the department: here. Students with other exam regulations can attend the course, but cannot do the exam.