The Privacy & Sustainable Computing Lab together with GlobArt and Capital 300 hosted a Round Table discussion about artificial intelligence (AI), Ubiquitous Computing and the Question of Ethics on the 9th of November 2017 in Vienna. We were happy to have Jeffrey Sachs as our distinguished guest at this 4-hour intense Workshop on the future of AI. Other distinguished speakers were Bernhard Nessler from Johannes Kepler University Linz introducing to the limits of AI as well as Christopher Coenen unveiling the philosophical and historical roots of our desire to created artificial life.
The session and its speakers were structured by three main questions: What can general AI really do from a technical perspective?
What are the historical and philosophical roots of our desire for artificial life?
What sorts of ethical frameworks should AI adhere to?
The speakers argued that there is a need to differentiate between AI (Artificial Intelligence) and AGI (Artificial General Intelligence), where AI (like IBM Watson) needs quality training as well as quality data, lots of hardware and energy. In contrast, AGI is able to work with unstructured data and can have a better energy consumption rate. The other advantage of AGI is that it can react to un- foreseen situations and could be more easily applicable to various areas. One point that was stressed during the debate was that a lot of the terminology used in the scientific field of AI and AGI is borrowed from neuroscience and humans proper intelligence. Since machines – as experts confirmed – do not live up to this promise, using human-related terminology could lead to a misleading of the public as well as overly confident promises by industry.
It was discussed whether the term ”processing” might be more suitable than ”thinking” – at least at the current state.
Another phenomenon could be due to science fiction (Isaac Asimov, Neal Stephenson …) or Movies like ”Her” or ”Ex Machina”, where we rather should differentiate the terms AGI and Artificial Life.
What are the socio-cultural, historical and philosophical roots of our desire to create a general artificial intelligence and to diffuse our environments with IT systems?
”The World, the Flesh & the Devil” a book published in 1929 by J. Desmond Bernal was a named inspiration for the concept of the ”mechanical man”. This book in turn provided an excellent introduction into the debate about transhumanism, which often goes hand in hand with the discussion about AI. Some prominent figures in technology – such as Ray Kurzweil or Elon Musk – frequently communicate transhumanistic ideas or philosophies.
What ethical guidance can we use as investors, researchers and developers or use in technical standards to ensure that AI does not get out of control? Concerning this question, there was a general agreement on the need to have some basic standards or even regulations of upcoming AI technology. Providing one example of such standards, the IEEE is working on Ethical Aligned Design guidelines under the leading phrase “Advancing Technology for Humanity.” Here particular hope is put into P7000 (Model Process for Addressing Ethical Concerns During System Design) that sets out to describe value based engineering. Value based engineering is an approach aiming to maximize value potential and minimize value harms for human beings in IT-rich environments. The ultimate goal of value based engineering is human wellbeing.
In conclusion, the event provided an excellent basis for further discussions about AI and it’s ethics for both experts and students alike.
Speakers at the Roundtable:
- Christopher Coenen from the Institute for System Analysis and Technology Impact Assessments in Karlsruhe
- Peter Hampson from the University of Oxford
- Johannes Hoff from the University of London
- Peter Lasinger from Capital 300
- Konstantin Oppel from Xephor Solutions
- Michael Platzer from Mostly AI
- Bill Price who is a Resident Economist
- Jeffrey Sachs from Columbia University
- Robert Trappl from the Austrian Research Institute for AI
- Georg Franck who is Professor Emeritus for Spacial Information Systems
- Bernhard Nessler from Johannes Kepler University
- Sarah Spiekermann – Founder of the Privacy & Sustainable Computing Lab and Professor at WU Vienna.