„… also wir spielen da ein bisschen Gott…“ Soziale Robotik und die Ethnomethoden zur Herstellung von Menschen als soziale Akteure

Im Zentrum eines aktuellen Forschungsvorhabens steht die Eruierung einer ‚speziellen‘ Ethnomethodologie moderner Gesellschaften zur Herstellung von „Menschen“ als (einzige) legitime soziale Akteure. Insbesondere die Rolle der Technik- und Lebenswissenschaften eignen sich für eine empirische Untersuchung. In diesem Zusammenhang sind neben der Feldforschung in Deutschland auch einige Forschungsaufenthalte in Japan geplant. Die in Japan erhobene Empirie soll komparatistisch-kontrastierend zu den in Deutschland erhobenen Daten ins Verhältnis gesetzt werden.

Das empirische Feld, an dem ich versuche die Ethnomethoden eines für moderne Gesellschaften typischen „manufacturing humans“ herauszuarbeiten, ist das der ‚social robotic‘, der ‚wearables‘ und so genannter ‚smart environments‘. Im Forschungsfeld der smart environments kreuzen sich die Aktivitäten und Zielsetzungen der anderen zwei Felder, da in einer ‚intelligenten‘ bzw. ‚algorithmisch durchsetzten‘ Umwelt die rein technischen Artefakte (Roboter) und die am Menschen angebrachten Sensoren (wearables) ausgelesen, ausgewertet und zueinander kompatibel gemacht werden.

Obschon die Grenzen zwischen dem organisch-körperlichen, dem intentional-geistigen und dem mechanisch-technischen sowie maschinenhaft-algorithmischen de facto ununterscheidbar ineinanderfließen, werden die Grenzen umso vehementer durch entsprechende Diskurse aufrechterhalten. Wie wird das genau bewerkstelligt (Praxis)? Welche wissenschaftsinhärenten Gründe könnte es hierfür geben (Epistemologie)? Was steht auf dem Spiel, wenn die diesbezüglichen Grenzen einstürzen (Politik/Macht)? Kurzum: Wer sind die Verlierer und wer die Gewinner einer kontrafaktischen Aufrechterhaltung einer Grenze zwischen ‚natürlichen‘ Menschen und ‚artifizieller‘ Technik? Und schließlich: Auf welch eigentümliche Weise sind in dieser Hinsicht wissenschaftliche Vorgehensweisen und gesellschaftliche Interessen ineinander verwoben ohne es voneinander zu wissen?

Kernenergie in Europa

Europa ist in Fragen der Kernenergienutzung gespalten. Nach Fukushima (2011) und dem Pariser Klimarahmenvertrag (2015) gilt dies unverändert fort.

Der Vortrag versucht einige wesentliche Aspekte der Lage und der Trends hinsichtlich der Kernenergie in Europa zu skizzieren. Dabei werden eine Reihe von Fragen berührt – insbesondere: Wie ist der nukleare Status-quo beschreibbar und einzuschätzen? Welche Ambivalenzen und Risiken sind mit der Kernenergienutzung verbunden? Sind die mit der Kernenergie verbundenen Versprechungen erfüllbar? Wie sieht es am Front-end (Uran) und am Back-end (Abfall) aus? Welche Zukunftstechnologien werden angestrebt? Gibt es europäische Forschungsprioritäten? Wer agiert? Was bedeutet das für die Technikfolgenabschätzung?

Hermeneutik der Zukunft. Verstehende Erschließung der künftigen Technologien

Der Vortrag beschäftigt sich mit der Frage nach der hermeneutischen Zugangsweise zum Thema Zukunftstechnologien. Dabei gilt es zunächst zu verdeutlichen, in welchem Sinne hier von der Hermeneutik der Zukunft gesprochen wird. Einerseits geht es um die These vom konstitutiven Vorrang der Zukunftsekstase für die Begründung der Hermeneutik (im Anschluss an Heidegger). Andererseits wird die Zukunft als das zu Verstehende bzw. das Objekt der hermeneutischen Verfahrensweise dargelegt. Im Zuge der Klärung dieser Fragen soll auch auf die Vielfalt der hermeneutischen Ansätze eingegangen werden, nicht zuletzt um auf die in ihnen vorzufindende Heterogenität bezüglich der Problemfelder Geschichte, Kommunikation, Kontext, Narration (Schleiermacher, Dilthey, Gadamer, Kritische Theorie, Objektive Hermeneutik) hinzuweisen. Die Bemühung der Hermeneutik bzw. des verstehenden Zugangs zu der Frage nach den technisch geprägten Zukünften soll diese Vielfalt in Betracht ziehen.

Realising the policy potential of history – challenges and opportunities from the History & Policy project

As complexity and uncertainty increase, the use of scenarios to exploring that uncertainty becomes essential to support decision makers. Yet, to respond to that need, a paradigm shift from traditional, “paper and pencil” approaches (albeit supported by complex software) toward simulation-based approaches is needed, not least because the cognitive demand required to envisage all possible alternatives a business/political decision might entail, has become too high to manage for the human mind. The practical need is both to unburden analysts from the cumbersome task of manually deriving scenarios and allow for scenarios to be properly managed. In this talk, starting from a preliminary introduction on the importance of using the appropriate reasoning process (inductive vs. deductive reasoning), it will be presented how the Holistic Risk Analysis and Modelling (HoRAM) method (patent pending), by leveraging on (logic-based) artificial intelligence, allows both to generate all possible scenarios associated with a decision and to calculate and represent the risk profile associated to them, even without the need of statistically significant data (i.e., even with qualitative data elicited through interviews). Further, HoRAM, which is applied through the cloud-based Klarisk® platform, allows to identify and prioritize the criticalities associated with the analyzed decision on the basis of their contribution to the overall risk and, most of all, to compare and verify (before they are applied) the goodness of the different alternatives, thus increasing the capability to search for the best risk strategies.

Wissenschaft und Politik: Lernprozesse durch das gemeinsame Explorieren von Politikalternativen

Welche Rolle sollte wissenschaftliche Expertise an der Schnittstelle zu Politikprozessen einnehmen? Diese alte Frage wird hier in aktuelle Kontexte der Klima- und Nachhaltigkeitspolitik gestellt mit einem Fokus auf formalisierte, globale wissenschaftliche Assessment-Prozesse zu Umweltfragen. Die langjährigen, politisch beauftragten Assessmentprozesse wie die des Weltklimarates (IPCC) bewerten den wissenschaftlichen Kenntnisstand zu einem bestimmten politischen Themenfeld. Insbesondere mit der normativen Komponente der Bewertung von Politikoptionen haben viele Assessmentprozesse noch keinen guten Umgang gefunden. Ein neues Modell wird hier vorgeschlagen. Die Annahme einer Ziel-Mittel-Interdependenz vermittelt über die praktischen Konsequenzen von Politikpfaden, die Erforschung von Alternativen, sowie der öffentliche, inklusive und deliberative Diskurs zu diesen Fragen sind die wesentlichen Bausteine des pragmatischaufgeklärten Modells (PEM), das hier theoretisch und praktisch vorgestellt wird und im IPCC bereits Anwendung gefunden hat. Es ist „aufgeklärt“, weil es über die praktischen Konsequenzen der Mittel informiert, die zur Erreichung politischer Ziele nötig sind. Positive oder negative Nebenwirkungen alternativer Pfade zeigen, dass in den Entscheidungen implizit meistens noch andere Ziele und Werte zu berücksichtigen sind. Die Kartografie alternativer Politikpfade kann deliberative Lernprozesse aller Beteiligten ermöglichen, was den populistischen Tendenzen entgegengehalten werden kann.

Futures of Science and Technology in Society

In his famous Rede Lecture on “Two Cultures and the Scientific Revolution” in 1959, the UK scientist and science adviser (and novelist) C.P. Snow noted that “scientists have the future in their bones”. This can be read as a suggestion of an occupational disease of scientists; but of course only in a metaphorical way. It does imply a diagnosis of scientists (and for that matter, of engineers) in our modern society, and their penchant to paint a cornucopian future if only they were given the means (and the space) to realize their promises. This is still visible in our post-modern (or at least reflexively modern) times, even if there are now also contrary voices doubting some of the promises. Think of the critical movements, partly within science, since the 1960s, and the emergence of technology assessment as an institutionalized function of governments - and of society generally.

This is only part of the story that I want to develop about futures of science and technology in society, and which is visible in the collection of published articles of mine that is the occasion for this seminar. But it still starts with a diagnosis, about the nature of doing modern science, and the path dependencies involved; one example of course being paradigms in the sense of Thomas Kuhn, and the “normal science” that ensues - at least for a time.

The focus of my diagnosis is not about the struggle between promises, and between promises and potential concerns, however. It is about the patterns that develop and stabilize, including the institutions of science and technology as we have come to know them. These are long-term developments, as I have tried to show in my diagnosis of how science lives, and has to live, in protected spaces, at the micro-level in laboratories and field experiments; at the meso-level in the world of scientific disciplines and funding bodies of various kinds; and at the macro-level of justification/legitimation of science in the wider world, still relatively stable in spite of occasional contestation from various sources.

This diagnosis can and should lead to reflection: what happens if the protected spaces lose their protection? Or are disturbed, for internal or external reasons? One reaction, call it “conservative” for lack of a better term, is for scientists and spokespersons for science to try and keep protected, and come up with justifications for it (and there might well be good reasons). The opposite reaction is to embrace the disturbances and disorder, as in the creativity of “anything goes” (Feyerabend) or “creative destruction” (Schumpeter, and present-day adepts of “disruptive innovation”). I have argued elsewhere for a “third way”, of learning how to live in unprotected spaces.

I have tried out my hand at actually delineating possible futures of science and technology in society. And developed an approach to such exercises; a “foresight methodology” if one wishes to use that term. Starting with the futures that are embedded in present patterns and institutions, “endogenous futures” as it were, but keeping visible the tensions and struggles that are part of it. The latter may be articulated further in scenarios in which the endogenous future might collapse, partly for external reasons but also because of internal dynamics. I offer a tongue-in-cheek, but serious, example in the article offering a scenario of the collapse of the struggle for excellence. We have also explored such possibilities in sociotechnical scenarios for the development and societal embedding of nanotechnology, nuancing the cornucopian promises. The same sort of dual approach can be used for the presently (still) fashionable program on Responsible Research and Innovation (the term is used for a cross-cutting theme in the EU Framework Programme Horizon 2020), which I have analysed as a social innovation, and still open-ended, but with emerging path dependencies.

This “foresight methodology” is informed by a sense of social order as a precarious achievement, to be cherished but not take as natural and self-justificatory either. In other words, the ambivalence of an order that is socially constructed, and in that sense contingent; but may have good reasons to be maintained.

This is a much more general point than futures of science and technology in society. It links up with the tragic notion of Gesellschaft, die ärgerliche Tatsache. So it is fitting to go broader and deeper in the final two papers in the collection. One is about technology as a material narrative, and the prospective ontology it implies. Engineers have the future in their bones, and so has technology. The other, as yet unpublished paper, emphasizes the granularity of social order, and how complementary repertoires (and the grey zones that go with them) constitute order – precariously.

Here this speaker for the futures of science and technology rests his case.

Autonomous Systems and the Value Alignment Problem

Several thinkers have suggested that we should impose our own values and norms on autonomous systems (such as self-driving cars) to ensure they serve human needs and wishes. Stuart Russell calls this the value alignment problem: How can we build autonomous systems with values that “are aligned with those of the human race”? As he sees it, the challenge is to build autonomous systems with ethical priorities that do not pose a threat to us. He argues that the machine’s purpose should be to “maximize the realization of human values.”
The IEEE, the world’s largest professional organization for engineers, has issued a report on the value alignment problem entitled Ethically Aligned Design: A Vision for Prioritizing Human Well-being with Autonomous and Intelligent Systems (A/IS). This document is part of the IEEE Global Initiative on Ethics of Autonomous and Intelligent Systems. The authors claim that autonomous systems “should always be subordinate to human judgment and control. […] If machines engage in human communities as autonomous agents, then those agents will be expected to follow the community’s social and moral norms.”
In this talk I will defend two claims about the value alignment problem: (i) It is not obvious that it is desirable to build autonomous systems with values that “are aligned with those of the human race”. This is a substantial moral claim that needs to be critically discussed and supported with arguments. (ii) The methods currently applied by computer scientists for embedding moral values in autonomous systems can be improved by representing moral principles as conceptual spaces, i.e. as Voronoi tessellations of morally similar choice situations located in a multidimensional geometric space. It is a mistake to use utility functions.
The first point will not surprise philosophers, but I think it is worth stating clearly. The second point, which is perhaps the most interesting one, requires some unpacking before it can be assessed and discussed. In the talk I will argue that it would be a mistake to always align the values of autonomous systems with those embraced by human beings, because humans are sometimes wrong about what is valuable. I will also discuss some links between general ethical theories and the value alignment problems. Because experts disagree on which ethical theory is correct, it would be ill-advised to base a solution of the value alignment problem on any ethical theory. My argument for this claim is closely intertwined with my suggestion for how to represent moral principles in autonomous systems: The point of departure for this proposal is some ideas developed by two leading cognitive scientists, Eleanor Rosch and Peter Gärdenfors. I will try to show that some of their insights may be applicable to the ethics of autonomous systems.

Towards a value-sensitive design of energy transitions. Technology and human wellbeing

Central aim of this paper is to argue that technologic innovations are coupled to our concept of the good life, our concept of (human) wellbeing. I contend that technologic progress has to be guided by this ideal. The argument will be developed by applying the concept of value-sensitive design to large-scale technological systems like the energy system. These are in large parts driven by political decisions and affect us all, making an inter-subjective concept of assessment necessary. I want to depict such a concept of wellbeing by applying the capability approach to the energy transition and argue that the capability approach provides a normative framework that can be seen as an individualistic foundation for sustainability analysis in the energy context.

The Global Consortium for Sustainability Outcomes (GCSO): Design, Status, Prospects

The Benefits of Public Engagement?

The idea of involving the public in aiding policy decisions (such as in areas of science and technology) has become popular in many democratic societies over recent years. However, the benefits of such ‘public engagement’ have not been convincingly demonstrated. In this presentation I will argue that public engagement is not an intrinsically good phenomenon, and that its proper use, formulation, and implementation should be informed by evidence as provided by ‘evaluations’. I will discuss the rationale for, and difficulties of, evaluation. A number of evaluation examples will be provided to demonstrate the kinds of processes that may be employed in such assessments and their limitations. I will conclude with a discussion of a number of dilemmas in evaluating engagement activities that, together, have worrying consequences. Finally, a few thoughts on the future of public engagement will be shared.

Related article: https://doi.org/10.1080/01442872.2018.1451502

Entseelte Entscheidungen - Sollen Computer über Menschen entscheiden?

Seit den 1980ern scheint sich das Bild des irrationalen menschlichen Entscheiders immer mehr zu verdichten, während gleichzeitig die Maschine als scheinbar objektiver Entscheider an Bedeutung gewinnt.

Ich zeige am Beispiel von Rückfälligkeitsvorhersagealgorithmen, welche Schwierigkeiten überwunden werden müssen, um qualitativ hochwertige algorithmische Entscheidungssysteme zu entwickeln und welche gesellschaftlichen Entscheidungen dafür jeweils getroffen werden müssen.

What responsibility should we attribute to innovators in the light of responsible innovation?

Responsible innovation has been described as ‘on-going process of aligning research and innovation to the values, needs and expectations of society’ (Rome Declaration on RRI). The notion suggests that innovators have additional responsibilities (to society, stakeholders, users) than in ‘regular’ innovation. I ask the question what these responsibilities are and if we can reasonably attribute these responsibilities to innovators. To that end I distinguish between the process and product dimension of responsible innovation and between back-ward-looking (or retrospective) and forward-looking (or prospective) responsibilities.

This creates a matrix of four types of responsibility that could in principle be attributed to innovators. I then discuss two kinds of problems in attributing these responsibilities to innovators. The first is that it would often seem unreasonable to allocate certain responsibilities to innovators. Innovation is typically a process that involves a multiplicity of authors, in which there are long causal chains between innovation and the eventual social impacts and in which many things are uncertain. I argue that this may lead to the problem of many hands, i.e. situations in which nobody is (seemingly) responsible.

A second problem is that even in cases in which it would seem reasonable to attribute responsibility, there may be trade-offs between the different kinds of responsibility. There is for example evidence that attributing backward-looking responsibility for defects in products diminishes the willingness to learn about such defects and to take forward-looking responsibility.

After discussing these problems, I explore possible directions for dealing with them. In particular, I will argue that we need a more fine-grained account of responsibility that distinguishes more carefully between different kinds of backward-looking and forward-looking responsibility.

Second, I argue that in responsible innovation it is crucial to organize learning both about (possible) social impacts of innovations and about how to socially evaluate such impacts, and that the attribution of responsibility should be attuned to facilitate such learning processes.

The emergent practice of transformative science for sustainability in Luxembourg: Underlying concepts, methods and processes for changing material engagements

Leading thinkers in a variety of scientific fields have identified a range of obstacles that prevent our society from learning to cope better with the existential challenges of the Anthropocene.

These include the increasing fragmentation of disciplinary knowledge, prevailing management practices that are based on fallacious notions of prediction and control, and active avoidance of explicit deliberations on values and judgment of diverse life forms/cultures in our pluralist societies, itself due to prevailing social norms on ‘political correctness’. Self-reinforcing patterns of distribution in neo-liberal capitalist systems also play a role.

In this seminar, we will explore the further development of concepts and methods for transformative science for sustainability, such as collaborative systems mapping, scenarios and citizen science. In these considerations, we will draw on Science and Technology Studies (STS) and pragmatist philosophy of experience and the mind, in order to develop methods that stand a better chance to address these obstacles to learning. In particular, we will discuss the promise of designing processes for dialogic social learning, combined with a theoretical framework for the evaluation of (learning) outcomes and impacts built on the Material Engagement Theory (MET) that is emerging from the field of cognitive archaeology. MET asks ‘how things shape the mind’, starting from conceptions of mind as embodied, extended and distributed and also entangled with the material world (e.g. by considering how material things can serve as enactive signs). MET presents a welcome complement to more internalist perspectives on what it might take to transform consciousness for sustainability offered by phenomenology.

I hope that the seminar will offer the opportunity to debate if and how MET might be adapted as a framework for the evaluation of the practice of transformative science, with reference to two projects in Luxembourg that build on the same concepts and methods. One project is concerned with sustainable school development in collaboration with three schools and the national school development agency. The other project is concerned with sustainable water governance and land use in two river partnerships and at the national level. In our view Luxembourg and these two projects present a particularly interesting field for the practice and evaluation of transformative science, as there are only two governance levels – national and municipal – and paths to government are exceptionally short. Most key stakeholders and opinion leaders in the two fields of policy and practice welcome and have started to engage in the projects. Fostering and evaluating learning that is indicative of how we engage with the material world at the individual, organizational and systemic levels, and studying how learning at these levels might be related, may become possible in this setting.