Paweł Łupkowski

The aim of this paper is to present and discuss the issue of the adequacy of the Minimum Intelligent Signal Test (MIST) as an alternative to the Turing Test. MIST has been proposed by Chris McKinstry as a better alternative to Turing’s original idea. Two of the main claims about MIST are that (1) MIST questions exploit commonsense knowledge and as a result are expected to be easy to answer for human beings and difficult for computer programs; and that (2) the MIST design aims at eliminating the problem of the role of judges in the test. To discuss these design assumptions we will present Peter D. Turney’s PMI-IR algorithm which allows for MIST-type questions to be answered. We will also present and discuss the results of our own study aimed at the judge problem for MIST.

In September 2018, the ArgDiaP association, along with colleagues from Germany and the UK, organised one of the longest and most interdisciplinary series of events ever dedicated to argumentation - Warsaw Argumentation Week, WAW 2018. The eleven-day ‘week’ featured a five day graduate school on computational and linguistic perspectives on argumentation (3rd SSA school); five workshops: on systems and algorithms for formal argumentation (2nd SAFA), argumentation in relation to society (1st ArgSoc), philosophical approaches to argumentation (1st ArgPhil), legal argumentation (2ndMET-ARG) and argumentation in rhetoric (1st MET-RhET); and two conferences: on computational models of argumentation (7th COMMA conference) and on argumentation and corpus linguistics (16th ArgDiaP conference). WAW hosted twelve tutorials and eight invited talks as well as welcoming over 130 participants. All the conferences and workshops publish pre- or post-proceedings in the top journals and book series in the field.

Our aim is to model the behaviour of a cognitive agent trying to solve a complex problem by dividing it into sub-problems, but failing to solve some of these sub-problems. We use the powerful framework of erotetic search scenarios combined with Kleene’s strong three-valued logic. ESS, defined on the grounds of Inferential Erotetic Logic, has appeared to be a useful logical tool for modelling cognitive goal-directed processes. Using the logical tools of ESS and the three-valued logic, we will show how an agent could solve the initial problem despite the fact that the sub-problems remain unsolved. Thus our model not only indicates missing information but also specifies the contexts in which the problem-solving process may end in success despite the lack of information. We will also show that this model of problem solving may find use in an analysis of natural language dialogues.

In this paper, we consider how an agent may manage a questioning agenda in a situation where multiple information sources are available. We work within the framework of formal dialogue systems with the underpinning of Inferential Erotetic Logic. Firstly, we present the formal dialogue system DL(IEL)mult for managing multi-agent information retrieval. Then, we extend the proposed system so that it is capable of representing group and individual levels for the question decomposition process. We also propose two measures for evaluating information sources: their cooperativeness and success levels; next, we analyse how the choice of agents may influence the way in which a solution for a given problem is reached.

This paper aims to present and discuss an argumentation against the Turing test (TT), which we shall call the CCSC (Complete Conversation System Claim). Exemplary arguments of the CCSC type include Lem’s “Space Gramophone,” the “machine equipped with a dictionary” proposed by Shannon and McCarthy, Block’s “Aunt Bubbles,” and Searle’s “Chinese Room” argument. CCSC argumentation is constructed to show that the TT is not properly designed and, consequently, is not a good hallmark of intelligence. Based on the original TT rules reconstruction, I argue that CCSC-type argumentation seems to be aimed at a certain interpretation of the TT, which, as I demonstrate, commits the straw man fallacy. In light of the results presented by Łupkowski and Wiśniewski, I also discuss whether a complete conversation system is theoretically possible.