LOGICS OF INFORMATION AND ACTION
(objectives)
We live in an information network and in an exchange of opinions that is ubiquitous and constant – a net of epistemic acts that we exchange with other agents and affect what we end up believing and deciding. Working with information implies more and more that we face the social effects of this – and these are today faster and faster, and we get a glimpse of them in real time. However, the more agents we have involved, the harder to understand the dynamics of information release turn to be.
This course introduces a formal toolkit that helps in this enterprise. In particular, the course aims at securing: (1) the understanding of the problems of reasoning that can be triggered by the release of information; (2) the understanding of models that capture the dynamic effects of information release, and the conceptual problems they raise; (3) the problems connected to the representation of belief-merging and, in general, the relations between individual and collective notions of epistemic attitudes; (4) the understanding of the conditions at which consensus is possible, the role it can play, and the relation between the information release policies, the connection of the epistemic network, and the hierarchies and trust distribution in epistemic communities.
(3) e (4) presuppose (1) and (2). In turn, the last two objectives come with a view on the social impact that the information release policies have on a community of epistemic agents. The course employs a varied package of methods and tools, especially those from Epistemic Logic and Dynamic Epistemic Logic, but also, to a lesser extent, notions and methods from Judgement Aggregation and Network Epistemology, which the course will briefly introduce.
|
Code
|
20710706 |
Language
|
ITA |
Type of certificate
|
Profit certificate
|
Credits
|
6
|
Scientific Disciplinary Sector Code
|
M-FIL/02
|
Contact Hours
|
36
|
Type of Activity
|
Related or supplementary learning activities
|
Teacher
|
CIUNI ROBERTO
(syllabus)
In 2003, Nick Bostrom conjectured that we could be living in a computer simulation. Twenty years later, progress in AI (Artificial Intelligence) has refreshed the interest in the theoretical scenario envisaged by the conjecture. The very same progress is making increasingly hard to tell apart an AI software from a ‘natural’ (that is, non-artificial) intelligent agent, at least in online environments. In this course, we will talk about all this: AI, simulations, and how hard it could be, at least in principle, to tell them apart from ‘natural’ intelligent agents and reality, respectively. We will see that this is a particular variation of a question that has been repeatedly asked along human civilization, and that such a question is relevant in all those scenarios in which we can imagine a systematic indistinguishability between two mutually exclusive alternatives that are logically distinct from one another. Standard examples here involve dream and reality, illusion and reality, simulation and reality, natural intelligent agents and artificial intelligent agents. The course will focus on how this indistinguishability is connected to the exact information available to us, and to the fact that the indistinguishability can persist even if we increase our information. In particular, we will discuss the implications of these scenarios for the stability and the truth of our beliefs on the distinction between simulation (illusion, dream) and reality, on the one hand, and the distinction between natural intelligence and artificial intelligence, on the other. The course will be taught in English.
(reference books)
Chalmers D. (2022) Reality+. Virtual Worlds and the Problems of Philosophy, Norton, New York.
Descartes R. (2005) Discourse on Method and The Meditations, Penguin, London. (Discourse on Method first published in French in 1637).
|
Dates of beginning and end of teaching activities
|
From to |
Attendance
|
not mandatory
|
|
|