ConsciousCoders.talk: AI & Big Data (March 3, 2018)
How does the rapid development of AI affect society, and our lives? How can we deal with challenges linked to AI and Big Data? What is the role of developers?
The idea Algorithms and Data. • new business segments • opportunities to solve problems • limitations and risks With opportunities, challenges and risks emerge.
The idea Society and Decision-makers. • living in a fast-moving world disrupted by innovations • no experience in novel technologies • facing decision towards an uncertain future How to cope with complex technologies that influence the way we live? With opportunities, challenges and risks emerge.
The idea Developers. • creating novel technologies • solutions for challenges of Society What impact does our work have and what is our role? How to cope with complex technologies that influence the way we live? With opportunities, challenges and risks emerge.
The idea What impact does our work have and what is our role? How to cope with complex technologies that influence the way we live? With opportunities, challenges and risks emerge.
idea What impact does our work have and what is our role? How to cope with complex technologies that influence the way we live? With opportunities, challenges and risks emerge. coordinate Developers for discussion and aggregate opinions enable Society to understand and cope with new technology shape development towards beneficial use of Technology Our Goals
• explore approaches through deep dives CONSCIOUSCODERS.workshop Conscious Coders | The formats CONSCIOUSCODERS.talk CONSCIOUSCODERS.publication CONSCIOUSCODERS.blog • present findings to society • spread the word through public discussion • provide inspiring ideas and developments • share best practices • make outcomes accessible • basis for further discussion
• explore approaches through deep dives CONSCIOUSCODERS.workshop Conscious Coders | The formats CONSCIOUSCODERS.talk CONSCIOUSCODERS.publication CONSCIOUSCODERS.blog • present findings to society • spread the word through public discussion • provide inspiring ideas and developments • share best practices • make outcomes accessible • basis for further discussion
question Recommendation Accountability - Clearly assign legal and moral responsibilities ❖ Emotional satisfaction ❖ Detering developers from being careless Explainability - Institutionalize reviews ❖ cf. DIN norm ❖ cf. GDPR Issue Acceptability Accountability Question of guilt Explainability ≠Transparency Background Which ratio needs which norms? AI Human Decision-making
by independent institutions - training data - factor decomposition • Legal framework - connect AI and law communities • Raise awareness • Notify about AI use Importance of transparency Cost autonomous driving credit score Netflix face recognition kernel parameter clustering components Transparency of AI – When do we need it? How can we implement it?
Balance of Power in Data-driven Societies? Issue Asymmetric distribution of information related to data → China’s citizen score in Europe as well? Recommendation Technological Solution: • Privacy assistance tools • Personalized privacy settings • Privacy policy analysis • Predictive data value analysis Political Solution: • Mandatory data mining audit (private companies, government) • Reporting on user specific data use (connected to privacy assistance tools) → Combining technological and political measures to achieve more transparency about data collection, distribution and processing → Empower individuals to achieve balance
be used to avoid dual usage? Challenges • Global common ground for moral values • Standards • Misuse of code • Unbalanced power Recommendation • Raising awareness • Teaching • Embracing existing standards (IEEE) • Blockchain for verification
AI actions/decisions understandable for humans? - Humans the weak part of the relationship? Behaviour: - Emotional/ natural behaviour -> how human like? - AI responsibility? Resources: - New job requirements? Possible Recommendations Interaction: - provide explanations for actions/ decisions - segregated environment and collaborations Behaviour: - Measuring scale, Test Scenarios, expert opinions - Legal and society Resources: - new interdisciplinarity Humans and machines working together – How are our workplaces going to change
Klaus Mainzer Founding Director of Munich Center for Technology in Society (MCTS) John Grant Director of Privacy and Civil Liberties Engineering at Palantir Technologies Alexander Waldmann Operative Director of appliedAI Initiative at UnternehmerTUM Gigo Board Member of muCCC, the Munich branch of Chaos Computer Club (CCC) Moderation: Markus Müller, ConsciousCoders Co-Founder