DIHB045 Етика и елегантност на изкуствения интелект (AI Ethics and Elegance )

Анотация:

The course acquaints students with modern socio-ethical problems of rapidly evolving internet technologies based on artificial intelligence (AI) systems, machine learning and other approaches to decision making by intelligent machines. It highlights the role of ethics in the development strategies and applications of AI systems through man-in-the-loop solutions. Students are expected to discuss historically developed standards in ethics and aesthetics as ‘building blocks’ in the AI fine-tuning of an ever-changing world. The course content is tied to the IEEE Global Initiative 'Ethics of Autonomous and Intelligent Systems – The Classical Foundations' through the prism of current multidisciplinary trends.

From a methodological point of view, the course is intended to proceed as an associatively led practical discussion of multimedia presentations, focusing on key figures and concepts in the field of AI. This will sensitize their capacity to intuitively employ relevant skills based on good AI practices in their professional life. The awareness of social roles and communication techniques, as well as the selective operation of data sets on ‘predictable scenarios’ play a key role in the competitive positioning of students on the global markets.

During the learning process, lecturers communicate with students using the following social channels:

• Zoom, WhatsApp, Viber, WeChat (for videoconferencing of students)

• Moodle (for lecture materials and case studies)

• Facebook groups such as facebook.com/groups/dhsofia (to discuss materials between students and lecturers)

• Google Docs (for processing and synchronizing drafts on joint projects)

• Dropbox, WeTransfer, Telegram (for sharing and synchronizing files up to 2 GB).

прочети още

Преподавател(и):

 Людмил Дуриданов  д-р
гл. ас. Христина Костадинова  д-р
проф. Симеон Симов  д-р

Описание на курса:

Компетенции:

Upon course completion students:

1) know:

 about different stages of AI development in the social environment of digital media and services

 about AI's ‘embedded ethical standards’ in digital services as "moral agents"

 that historically developed intrinsic connections between ethics, aesthetics and elegance are fundamental for AI employment

 typical cases and market practices of the social think tanks Apple, Google, Facebook, Amazon and IBM Watson.

2) are able to:

 successfully distinguish between different types of "moral agents" in digital services,

 operate successfully with key concepts in ethics and aesthetics in various fields of machine learning,

 operate with the ‘building blocks’ of already developed ethics and aesthetics.


Предварителни изисквания:
Students are expected to have the following skills:

o Working English

o Basic computer skills



Форми на провеждане:
Редовен

Учебни форми:
Лекция

Език, на който се води курса:
Английски

Теми, които се разглеждат в курса:

  1. Key multidisciplinary moments in the development and application of AI in various social fields (Predictive Analytics, Machine Learning, Social Data Science, Evolutionary Algorithms, Deep Learning, Robotics, Human-Machine Interaction).
  2. Key ethical standards in the 21st century, based on Kant's postulates (= to the 1789 French Social Revolution): 1) What can I know? (boundaries and transparency of knowledge -> Philosophy) 2) What should I do? (wisdom of free choice -> Moral, Ethics) 3) What may I hope for? (the heretical imperative of the 21st century -> Religion, Ritual and Belief) 4) What is man? (creator or mediator of environment -> Anthropology) 5) What is AI? (autonomous and creative functions of AI -> Digital Humanities / Philosophy, Psychology, Robotics)
  3. The three digital revolutions and their aesthetic and ethical impact - Digital isolation and creation of new social differences (revolution of the new social divide), - Dynamic inequality of access to real-time knowledge (the smart intelligence revolution), - comprehensive real-time control (the surveillance and control revolution).
  4. The four waves of AI development and distribution: Acceptability of algorithms to have autonomous development (since 1998 - Internet AI, since 2004 - Business AI, since 2011 - Perception AI, since 2015 - Autonomous AI)
  5. 21st century analytical ethics To what extent is it ethically acceptable to collect datasets (e.g. the 5000 features’ granulation during the Alamo / Cambridge Analytica project) and use them as an online tool for favoring real-time solutions to each mobile user via personalized messages and purchase offers as well as to impact political choices and crisis solutions such as the current pandemic (e.g. the COVID Safe app); algorithmic solutions of the massive surveillance by government institutions and global players (after Edward Snowden's disclosures)? The customer intimacy of massive natural language AI apps.
  6. AI ethics and digital power How to embed ethically compliant deep fake filters without creating a ‘digital dictatorship’, such as patterns and clusters in the development of a successful business and democratic policy (e.g., during election campaigns, such as those of Barack Obama, Francois Hollande, Donald Trump or voting pro / contra BREXIT).
  7. Ethics and aesthetics - two sides of the same coin The aesthetic education of a citizen (Friedrich Schiller) and Kant’s aesthetics (= to the 1789 French Revolution); how can elegance be objectively associated with ethics of free choice?
  8. 21st century elegance – the leading role of design (functionality, attractiveness, and transparency of AI) „Everyone should wear the dress that matches one’s essence“ (Maggy Rouff, La philosophie de l‘elegance); what dress suits AI the best?
  9. Elegance in a hybrid society - a criterion for social intelligence If each era has its own style of elegance and individuals „have their own fate“, would it be necessary for digital assistants to „have a destiny of a never aging child”, just being "invisible" functional robots?
  10. The social acceptance of "humanized robots" in a hybrid society Should AI robotics acquire the beauty and elegance of human behavior? How to bridge the 'uncanny valley', i.e., the annoying feeling that we are "duplicated".
  11. Aesthetics of an "ageless AI" vs human wisdom How would an elegant "ageless AI" which collects only knowledge experience, but not human wisdom to choose or give up (even having the opportunity to choose otherwise) look like?
  12. Ethics of intelligent machines In which AI cases would we take a ‘middle-aged individual’ as a template for the four types of machine agents: ethical agents (e.g., robot jockey), implicit ethical agents (e.g., safe autopilot), explicit ethical agents (e.g., using formal utility assessment methods) and full-fledged ethical agents (who can make explicit ethical judgments and can generally reasonably justify them)? What ethical standards are to be embedded to solve contradictions based on common sense (e.g. autopilots to choose how to make a car accident - with how many and what kind of victims and damages)?
  13. AI design – ethical recommendations In which cases can different computerized recommendation methods (e.g., choice architects and collaborative filtering) assist in the selection of elegant scenarios and ethical models of asynchronous communication to awaken passive or active interest in the ‘long journey’ with clients and partners, which should not be related only to aesthetic criteria.
  14. Ethics of machine decisions and freedom of human choice How can predictive analytics assist in making ethical and moral legal decisions, e.g., in the field of crime predictability and police intervention by creating psychological profiles (from IBM Bluemix to Minority report); selection of job positions (AI job interview, e.g. hirevue.com); health care robots.
  15. Development and application of algorithmic ethics Developing ethical standards for designing the behavior of intelligent machines (e.g., designing stand-alone solutions for digital assistants).

Литература по темите:

AI IEEE Global Initiative on Ethics of Autonomous and Intelligent Systems, Version 2., April 2019.

AI Ethics Principles, Australian Government, Department of Industry, Science, Energy and Resources

ANDERSON, MICHAEL & ANDERSON, SUSAN: Machine Ethics, Cambridge University Press 2011.

ASARO, PETER: AI Ethics in Predictive Policing. From Models of Threat to an Ethics of Care. In: IEEE Technology and Society Magazine, June 2019, pp. 41-53.

BROUSSARD, MEREDITH: Artificial unintelligence. How computers misunderstand the world, Cambridge: MIT Press 2018.

ДЕНКОВА, ЛИДИЯ (съст.): За нормата и шестото чувство. Естетически есета върху вкуса, София: Издателство на НБУ 2017. (добра антология за етиката и естетиката на вкуса преди Френската еволюция, повлияла Кант)

DOURISH, PAUL: The Social Life of Algorithms, Melbourne University: Pursuit 2016.

DURIDANOV, LUDMIL / SIMOFF, SIMEON: AI Ethics and Elegance – Just Married? – CEUR Workshop Proceedings (Т. 2413, Selected Papers of the XXII International Conference Enterprise Engineering and Knowledge Management (EEKM 2019), Moscow, April 24-26, 2019), pp. 15-22.

ELLIOT, ANTHONY: The Culture of AI. Everyday Life and the Digital Revolution, London & New York: Routledge 2018.

GALLOWAY, SCOTT: The Four – the Hidden DNA of Amazon, Apple, Facebook and Google, London: Penguin, Transworld Publishers 2017.

GORDON, THEODORE & TODOROVA, MARIANA: Future Studies and Counterfactual Analysis.Seeds of the Future, Springer / Palgrave Macmillan 2019.

IPHOFEN, RON & KRITIKOS, MIHALIS: Regulating artificial intelligence and robotics - ethics by design in a digital society, In: Contemporary Social Science 2019, pp.1-19.

JENNINGS, CHARLES: Artificial Intelligence. Rise of the Lightspeed Learners, London: Rowman & Littlefield Publishers 2019

KANT, IMMANUEL: Immanuel Kants Logik, ein Handbuch zu Vorlesungen, herausgegeben von Gottlob Jäsche, Königsberg 1800. In: Kant Werke, Band 5. Schriften zur Metaphysik und Logik, 1958, repr. 1983, pp. 421-582.

KANT, IMMANUEL: Critique of the Power of Judgment (Kritik der Urteilskraft), translation by Paul Guyer & Eric Matthews, Cambridge University Press 1970, repr. 2000.

KEARNS, MICHAEL & ROTH; AARON, The Ethical Algorithm, The Science of Socially Aware Algorithm Design, Oxford University Press 2020.

KITTLER, FRIEDRICH: Draculas Vermächtnis. Technische Schriften, Leipzig: Reclam Verlag 1993.

KUCKLICK, CHRISTOPH: Die granulare Gesellschaft. Wie das Digitale unsere Wirklichkeit auflöst, Berlin: Ullstein 2016.

MÜLLER; VINCENT: Ethics of Artificial Intelligence and Robotics. – Stanford Encyclopedia of Philosophy, 2020.

НИКОЛОВА, ЦВЕТОМИРА: Хипотезата за „странната долина“ и анимацията, София: Издателство на Нов български университет 2014.

PEREIRA, LUIS & LOPES, ANTONIO: Machine Ethics. From Machine Morals to the Machinery of Morality, Cham: Springer 2020.

ROUFF, MAGGY: La philosophie de l’élégance, Paris 1942.

SCHILLER, FRIEDRICH: On the Aesthetic Education of Man in a Series of Letters, translated with an introduction by Snell, R. New Haven, Connecticut: Yale University Press 1954.

SEEMANN, MICHAEL: Das Neue Spiel. Strategien für die Welt nach dem digitalen Kontrollverlust, Orange Press 2014.

SEEMANN, MICHAEL: Digital Tailspin. Ten Rules for the Internet After Snowden, Amsterdam: Institute of Network Cultures 2015.

SNOWDEN, EDWARD: Permanent Record, London: Pan Macmillan 2019.

STRONG, COLIN: Humanizing Big Data. Marketing at the Meeting of Data, Social Science and Consumer Insight, London: Kogan Page 2015.

TEGMARK, MAX: Life 3.0. Being Human in the Age of Artificial Intelligence, New York: Penguin, Alfred Knopf 2017.

ТОДОРОВА, МАРИАНА: Изкуственият интелект. София: Изток-Запад 2020.