Thomas Metzinger

Die Übernahme des KI-Unternehmens DeepMind durch Google für rund eine halbe Milliarde US-Dollar signalisierte vor einem Jahr, dass von der KI-Forschung vielversprechende Ergebnisse erwartet werden. Spätestens seit bekannte Wissenschaftler wie Stephen Hawking und Unternehmer wie Elon Musk oder Bill Gates davor warnen, dass künstliche Intelligenz eine Bedrohung für die Menschheit darstellt, schlägt das KI-Thema hohe Wellen. Die Stiftung für Effektiven Altruismus (EAS, vormals GBS Schweiz) hat mit der Unterstützung von Experten/innen aus Informatik und KI ein umfassendes Diskussionspapier zu den Chancen und Risiken der künstlichen Intelligenz verfasst und geht darin auf aktuelle, mittel- und langfristige Herausforderungen im Bereich der KI-Entwicklung ein.

This is an edited collection of 39 original papers and as many commentaries and replies. The target papers and replies were written by senior members of the MIND Group, while all commentaries were written by junior group members. All papers and commentaries have undergone a rigorous process of anonymous peer review, during which the junior members of the MIND Group acted as reviewers. The final versions of all the target articles, commentaries and replies have undergone additional editorial review. -/- Besides offering a cross-section of ongoing, cutting-edge research in philosophy and cognitive science, this collection is also intended to be a free electronic resource for teaching. It therefore also contains a selection of online supporting materials, pointers to video and audio files and to additional free material supplied by the 92 authors represented in this volume. We will add more multimedia material, a searchable literature database, and tools to work with the online version in the future. All contributions to this collection are strictly open access. They can be downloaded, printed, and non-commercially reproduced by anyone. DOI: 10.15502/9783958571020 http://dx.doi.org/10.15502/9783958571020 ISBN: 978-3-95857-102-0

conscious content like ``the self in the act of In 1989 the philosopher Colin McGinn asked the knowing'' (see, e.g., chapters 7 and 20 in this following question: ``How can technicolor phe- volume) or high-level phenomenal properties like nomenology arise from soggy gray matter?'' ``coherence'' or ``holism'' (e.g., chapters 8 and 9 (1989: 349). Since then many authors in the ®eld in this volume). But what, precisely, does it mean of consciousness research have quoted this ques- that conscious experience has a ``content''? Is tion over and over, like a slogan that in a nut- this an entity open to empirical research pro- shell conveys a deep and important theoretical grams and interdisciplinary cooperation? And problem. It seems that almost none of them dis- what would it mean to map this content onto covered the subtle trap inherent in this question. physical states ``under a certain description''? In The brain is not gray. The brain is colorless. other words: What kinds of relations a

This metatheoretical paper investigates mind wandering from the perspective of philosophy of mind. It has two central claims. The first is that, on a conceptual level, mind wandering can be fruitfully described as a specific form of mental autonomy loss. The second is that, given empirical constraints, most of what we call “conscious thought” is better analyzed as a subpersonal process that more often than not lacks crucial properties traditionally taken to be the hallmark of personal-level cognition - such as mental agency, explicit, consciously experienced goal-directedness, or availability for veto control. I claim that for roughly two thirds of our conscious life-time we do not possess mental autonomy (M-autonomy) in this sense. Empirical data from research on mind wandering and nocturnal dreaming clearly show that phenomenally represented cognitive processing is mostly an automatic, non-agentive process and that personal-level cognition is an exception rather than the rule. This raises an interesting new version of the mind-body problem: How is subpersonal cognition causally related to personal-level thought? More fine-grained phenomenological descriptions for what we called “conscious thought” in the past are needed, as well as a functional decomposition of umbrella terms like “mind wandering” into different target phenomena and a better understanding of the frequent dynamic transitions between spontaneous, task-unrelated thought and meta-awareness. In an attempt to lay some very first conceptual foundations for the now burgeoning field of research on mind wandering, the third section proposes two new criteria for individuating single episodes of mind-wandering, namely, the “self-representational blink” (SRB) and a sudden shift in the phenomenological “unit of identification” (UI). I close by specifying a list of potentially innovative research goals that could serve to establish a stronger connection between mind wandering research and philosophy of mind.

I have learned a lot from Josh Weisberg’s substantial criticism in his well-crafted and systematic commentary. Unfortunately, I have to concede many of the points he intelligently makes. But I am also flattered by the way he ultimately uses his criticism to emphasize some of those aspects of the theory that can perhaps possibly count as exactly the core of my own genuine contribution to the problem—and nicely turns them back against myself. And I am certainly grateful for a whole range of helpful clarifications.

As Flanagan remarks at the outset, many philosophers and researchers in the cognitive and neurosciences today believe that a naturalistic solution to the mind-body problem will eventually be found. Optimistic attitudes of this sort are usually inspired by the remarkable theoretical success so far achieved under the information-processing approach. The information-processing approach rests on a number of ubiquitous background assumptions. The most central of these is that treating human beings and their brains as information-processing systems may open precisely those levels of description under which the classical philosophical questions can be reformulated and finally answered. Many philosophers today try to treat traditional puzzles in this field - like the mind-brain-relationship, intentionality or self-awareness - by analyzing them in a top-down-fashion, i.e. by reformulating them in an information-processing terminology. In doing so one of their goals, usually, is eventually to naturalize those theoretical problems, to transform them into problems tractable by empirical science.

In this speculative paper I would like to show how important the integration of mental content is for a theory of phenomenal consciousness. I will draw the reader's attention to two manifestations of this problem which already play a role in the empirical sciences concerned with consciousness: The binding problem and the superposition problem. In doing so I hope to be able to leave the welltrodden paths of the debate over consciousness. My main concern is to gain a fresh access to the familiar theoretical difficulties associated with the concept of 'consciousness'.

Cognitive enhancement aims at optimizing a specific class of information-processing functions: cognitive functions, physically realized by the human brain. This article deals with ethical issues in cognitive enhancement. It discusses some standard conceptual issues related to the notion of “cognitive enhancement” and then continues from a purely descriptive point of view by briefly reviewing some empirical aspects and sketching the current situation. Several enhancement strategies are being tested and used. Then the chapter offers some reflections on the treatment or enhancement distinction. It turns to normative issues by describing standard topics in current debates, then highlighting three examples of relevant novel questions under an ethical perspective. It is of central importance to be able to draw on long-term studies yielding data on the benefits, risks, and side-effects involved in the use of such substances over months and years. It concludes by making some general proposals for policy makers.

In Kenneth Einar Himma’s substantial commentary, there are a number of conceptual misunderstandings I want to get out of the way first. This will allow us to see the core of his contribution much clearer. On page 2, Himma writes about the problem of “explaining how it is that a particular phenomenal self is associated with a set of neurophysiological processes.” This philosophical question is ill posed: no one is identical to a particular phenomenal self. “Phenomenal self” must not be conflated with “me.” Under SMT, phenomenal selves, in standard situations, are highly specific forms of representational content. They are not particulars in an ontological sense. First, Himma introduces the notion of a “mental subject,” without giving any defining characteristics. He then proceeds to make a strong claim about conceptual necessity, presenting it as selfevident without an independent argument: “…it is not conceptually possible for a conscious mental state to occur that is not instantiated by a mental subject”. I must admit that I do not have this modal intuition, the point is not self-evident to me.

The representational dynamics of the brain is a subsymbolic process, and it has to be conceived as an "agent-free" type of dynamical self-organization. However, in generating a coherent internal world-model, the brain decomposes target space in a certain way. In doing so, it defines an "ontology": to have an ontology is to interpret a world. In this paper we argue that the brain, viewed as a representational system aimed at interpreting the world, possesses an ontology too. It decomposes target space in a way that exhibits certain invariances, which in turn are functionally significant. A challenge for empirical research is to determine which are the functional regularities guiding this decomposition process. What are the explicit and implicit assumptions about the structure of reality, which at the same time shape the causal profile of the brain's motor output and the representational deep structure of the conscious mind arising from it (its "phenomenal output")? How do they constrain high-level phenomena like conscious experience, the emergence of a first-person perspective, or social cognition? By reviewing a series of neuroscientific results, we focus on the contribution the motor system makes to this process. As it turns out, the motor system constructs goals, actions, and intending selves as basic constituents of the world it interprets. It does so by assigning a single, unified causal role to them. Empirical evidence now clearly shows how the brain actually codes movements and action goals in terms of multimodal representations of organism-object relations. Under a representationalist analysis, this process can be interpreted as an internal representation of the intentionality relation itself. We try to show how such a more complex form of representational content, once it is in place, can later function as the building block for social cognition and a for more complex, consciously experienced representation of the first-person perspective as well. The motor system may therefore play a decisive role in understanding how the functional ontology of the human brain could be gradually extended into the subjective and social domains.

The article presents a critical survey of the philosophical discussion of the mind-body-problem since the collapse of Rylean behaviourism. The major theories (identity theories, supervenience, emergentist materialism, dualist interactionism and functionalism) are sketched and briefly evaluated with regard to their advantages and disadvantages. The conclusion is that no satisfactory theory about the relation between mental and neurophysiological states exists today, but considerable progress has been made regarding the contours of this cluster of problems. A catalogue of criteria which every future theory concerning the mind-body problem must fulfill is given at the end.

differences between dreaming and waking consciousness as well. In this chapter, we will argue that these differences mainly concern the subjective quality of the dreaming experience. The interesting question, from a philosophical point of view, is not so much whether or not dreams are conscious experiences at all. Rather, one must ask in what sense dreams can be considered as conscious experiences, and what happens to the experiential subject during the dream state. Finally, in order to arrive at a more differentiated understanding of dream consciousness, we will contrast our analysis of ordinary dreams with lucid dreams, as well as with the varying degrees of lucidity and cognitive clarity seen in semi-lucid and prelucid dreams

Let me begin by focusing on the long list of agreements between the Dan Zahavi and me. As he is such a careful and scholarly author, there are almost no misunderstandings to get out of the way first. At the beginning of section 2, there is a conflation of different concepts of possibility. If we grant that imaginability is conceivability, if we pass over “practical” possibility as a non-defined term, and grant that by “physically” possible Zahavi very likely means “nomologically” possible, it still would present a major step to say that something is conceptually or metaphysically possible. Not everything that is logically or conceptually possible is metaphysically possible as well—metaphysically possible worlds have to be interpreted as a subset of logically possible worlds. How this can be done is a subject of intense and highly technical debates in current philosophy of mind. We cannot possibly enter this debate here, but let me just point out how, for instance, Zahavi’s remarks in the second paragraph of page 4 rest on a conflation between nomological and logical possibility.

Of all the current philosophical attempts to rescue the concept of “self” by working out a weaker version, one that does not imply an ontological substance or an individual in the metaphysical sense, Marcello Ghin’s is clearly my favorite. His reconstruction of the original theory is absolutely accurate and without any major misunderstandings. Enriching the concept of a “SMT-system” with the notions of “autocatalysis” and “self- sustainment,” and adding the intriguing idea that we are systems reflecting these processes on a new level of complexity, namely with the help of an integrated PSM on the level of conscious experience, seems the way to go if one wants to keep the concept of “self.” I have great difficulties in writing a reply to Ghin’s commentary, simply because I agree with so much in it. Let us see where his approach leads us