Programme timeline
29/09/2026
Conference Opening
17:00-17:10
Nature Writing | Machine Writing. On Growing and Generating in Digital Space
17:10-17:30
This talk explores the intersection of Nature Writing and Machine Writing in contemporary digital literature and media art, asking how generative practices engage with nature not merely as subject matter but as procedural logic — and what possibilities, tensions, and responsibilities emerge when algorithms do the writing.
No Goal, No Competition, No Problem: Play as a Primary Mode of Creative-AI Approach
17:30-17:50
AI is full of controversies. It inspires much hatred. Ecological catastrophe creating AI, AI as cultural theft at scale, zombie headless AI-deepfakes eroding democratic ideals, toxic AI diminishing cognitive skills, AI as treacherous sycophancy, AI as oligarchic corporate hype. The list is an ever-expanding recursive self-reinforcing litany of mistrust, malaise, devastation and despair.
Opposing opinions weave anxious threads: human obsolescence is feared to be imminent at the same time as AI seems like over-hyped corporate sales-mongering. Where can genuine play exist in such a contested domain? Can it exist? Is an imminent AI-apocalypse of automated warfare and robotic killing going to eradicate humanity? Will de-novo AI-engineered bio-weapons extinct everything? Will art and authentic human creative joy survive the increasing capacity of AI to emulate human genius? Possibly yes, possibly no. Yet, in the meantime, and as an inverse that is not intended to trivialize the profound problematic aspects of AI, it is possible as an artist to engage with the new capacities of AI with joy wonder acceptance love presence and revelation.
Play is so integrated into human neurophysiologies that even genocidal-apartheids, climate-catastrophes, oligarchic-inequity and a potential obliteration of all sapien utility in an AI-pocalypse cannot eradicate it. For the innocent aspect of beingness, AI is like a big box of new toys being replenished and upgraded every few months. Intricate subtle astonishing revelations emerge from tiny perturbations (prompts). AI epitomizes non-linear reward-succulent chaos. This presentation will examine the archetypal potentials of #genAI to offer pathways toward flourishing for all in spite of proliferating lethal and banal pathways
Playing with Identities: Who Speaks in AI Poetry Performance?
17:50-18:10
The paper explores the interplay between the lyric (text-internal) and the speaking (text-external) voices in performances of AI-generated poetry. Historically, the notion of the lyric voice has been strongly entangled with that of authorship, referring both to the poet’s unique style and to their authorial presence in the text. It is doubly associated with subjecthood – by virtue of being a voice (even if only in a metaphorical sense) and as an expression of one’s aesthetic creativity, which contemporary culture considers essential to being a human person.
As it inevitably articulates a kind of lyric voice, AI-generated poetry thus represent in the purest form the ideology of AI anthropomorphism – the conception of AI as the imitation of the human subject. However, while in a poetry reading by a human poet, the speaking voice embodies the lyric voice, establishing a continuity of the speaking subject, the opposite takes place in the performances of AI poetry.
Generated texts may be performed either by synthetic voices, which do not carry the claim to authenticity, or by human voices that – by virtue of explicitly not being AIs – emphasise the gap between the speaking and the lyric voices. This disruptive quality of voicing AI texts makes it a powerful poetic device to playfully deconstruct AI anthropomorphism – challenging human anthropomorphism in the process as well.
Discussion with Zuzana Husárová
18:10-19:00
Otevření výstavy Cesta básní
19:00-19:30
Interaktívna digitálna inštalácia, kde vaše kroky sú slovami na ceste básní. Ľudskí autori a neurónové siete sú vašimi sprievodcami. A cieľom je radosť z hravej poetickej tvorby. Dielo vzniklo ako súčasť TAČR grantu s využitím generatívneho LLM modelu EduPo. TAČR grant prepája Ústav pro aplikovanou lingvistiku UK ČR, ÚČL AV ČR, Didaktikon a VŠVU v Bratislave.
Uzavření projektu TAČR EduPo
19:30-19:45
Projekt EduPo založený na aplikování možností současné generativní umělecké inteligence pro potřeby literárního vzdělávání vstoupil do své finální fáze, a tak je již možné představit jeho hlavní výsledky. Je to především vzdělávací online aplikace, určená pro obohacení výuky poezie na základních a středních školách, kterou tímto zveřejňujeme a dáváme volně k dispozici. Aplikace EduPo umožňuje automatické generování české poezie pomocí specializovaných jazykových modelů. Je navržena jako pomůcka pro žáky a učitele umožňující interaktivní práci s básnickým textem a experimentování s řadou jeho stavebních kamenů. Aplikaci doprovází metodika včetně sady pracovních listů. Na práci týchž jazykových modelů je založen rovněž interaktivní didaktický exponát "Cesta básní" instalovaný v centru Didaktikon a vzdělávací workshopy věnované umělé inteligenci a poezii, které ho budou doprovázet. K výsledkům projektu patří rovněž Korpus české poezie 1780-2020, který představuje datovou základu, jež umožnila trénování modelů a bude nadále sloužit vědeckým účelům.
Na realizaci tohoto projektu (finančně podpořeného Technologickou agenturou ČR) se v letech 2023-2026 podíleli týmy ÚČL AV ČR, VŠVU, Didaktikon UK a ÚFAL MFF UK.
Představení metodiky EduPo
19:45-20:00
Proč učit stroje básnit?
Příspěvek se zaměří na přínosy aplikace specializované na generování poezie (EduPo) ve školní oblasti i v kontextu podobně zaměřených projektů v zahraničí. Na několika výběrových sondách tak představí komplexnější metodiku, která se teoretickému zakotvení i praktickému uplatnění generativní umělé inteligence ve výuce poezie věnuje.
30/09/2026
Otevření doktorandského bloku
15:00-15:10
Agentic Entity and Human Centered AI in an Anticipatory and Speculative Model of Future AI Art
15:10-15:30
The paper examines the relationship between the concept of the Agentic Entity and the framework of Human Centered AI within an anticipatory and speculative model of future AI Art. It is based on the assumption that current artificial intelligence systems cannot yet be understood as autonomous artistic subjects in a strong ontological sense. However, their growing capacity for adaptation, selection, generation, memory continuity and interaction suggests the emergence of new forms of distributed agency.
The Agentic Entity is therefore understood as a functional and relational model of emergent agency that arises within the loop of the human, the AI system and the environment. Human Centered AI represents a transitional horizon between today’s generative AI and possible future forms of artificial general intelligence. In this framework, the issue is not only the ethical control of technology by humans, but also the design of protocols through which cooperation between human intention and increasingly autonomous operations of the AI system can be tested.The anticipatory and speculative model of future AI Art does not aim to precisely predict the emergence of AGI or an artificial conscious entity. Rather, it models the conditions under which AI aesthetic processes could become more adaptive, meaning making and partially autonomous. AI Art thus appears as a laboratory for future forms of human artificial creativity, in which the understanding of the author, the creative process, aesthetic decision making and the relationship between human and nonhuman agency is transformed.
Artificial Intelligence in Digital Games
15:30-15:50
The conference paper Artificial Intelligence in Digital Games presents the research framework of a doctoral dissertation investigating the potential of digital games as environments for the acquisition, generation, and transformation of data used in artificial intelligence training. The research is based on the premise that digital game players constitute a unique source of information regarding human decision-making, problem-solving strategies, adaptation to novel situations, and social behavior. Through carefully designed game mechanics, such data can be systematically collected, analyzed, and subsequently utilized in machine learning processes.
The paper outlines the theoretical foundations of the dissertation, situated at the intersection of artificial intelligence, game studies, and human-computer interaction research. It further introduces the proposed research methodology, including the design of a digital game as a data collection environment, methods of data processing, and approaches to training machine learning models based on player-generated inputs. Preliminary findings from the ongoing experimental phase will be discussed alongside the challenges and limitations associated with transferring knowledge acquired in virtual environments to real-world applications.
The primary objective of the dissertation is to examine the hypothesis that digital game players can serve as effective sources of training data for artificial intelligence systems and that knowledge generated through gameplay can, under specific conditions, be applied beyond the boundaries of the game world. The research also explores the broader role of digital games as platforms for the development and training of future intelligent systems, extending their function beyond entertainment into the domain of AI research and innovation.
From Prompting to Bio-Actuation: EEG-Driven Shan Shui as a Playful Human–AI Language Game
16:10-16:30
This paper describes an artistic-research experiment that transforms EEG brain signals into AI-created Chinese Shan Shui art, defining the creative process as a playful human-machine "language game". Current generative AI communication often occurs through textual prompts, the proposed solution in this research is a transition from symbolic commands to embodied signaling. The human participant will become a bio-actuator in an inverse embodiment relation: (Machine−I) → World.
The EEG data are split up into time segments, and converted into feature vectors (alpha, beta, raw variability, signal quality). These features are correlated to visual input prompt parameters (e.g., mist density, ridge sharpness, compositional motion), through two separate workflows: API-based image generation, and SDXL-based live rendering. The generated results are compared with references from traditional ink landscape painting to assess both stylistic similarity and changes in process ontology.
The paper provides a further development of my dissertation framework, The Allegory of the Machine, that dataset-conditioned generation often takes place in a "Digital Cave" of probabilistic recombination. The study aims to explore how the EEG intervention introduces situated indeterminacy and material contingency to generate indeterminate outputs from generative models. In this connection, creativity arises from the co-play of neural signals, computational processes, and cultural visual tradition.
Discussion with Jana Horáková
16:30-17:00
Coffee break
17:00-17:20
Otevření bloku o hravosti v kontextu AI
17:20-17:30
Ve hře bez odstupu
17:30-18:00
V souvislosti s AI technologiemi se motiv hry objevuje hned v několika podobách. Jazykové modely se opírají o statistické vztahy mezi slovy, ale jazyková hra nikdy není pouze o jazyku, protože fundamentálně vzniká uvnitř forem života, kde má AI limity. Podstatnější problém spočívá v určité představě digitalizovaného světa, kterou AI institucionalizuje a v níž jsme nuceni existovat i hrát si. Herní principy jsou navíc klíčovou součástí designu aplikací, skrze něž se s technologií AI setkáváme, jde však spíše o formy manipulace a kontroly.
Ve svém příspěvku bych se ale ráda věnovala hře, jak ji chápe Alice Koubová, pro kterou představuje příležitost zaujmout tvořivý odstup od sebe sama, performativně experimentovat s identitou a vstupovat do rolí nezávisle na tom, kým jsme, čímž vzniká vzácný prostor pro reflexi, proměnu a také svobodu. Budu se ptát, zda interakce s AI otevírá prostor hry ve smyslu tohoto tvořivého odstupu, nebo zda nás naopak vede ke vstupování do rolí bez kýžené distance.
Z jeskyně do datového centra: interaktivní fikce, herní vyprávění a velké jazykové modely
18:00-18:30
Počátky textových her jsou svázány nejen s přírodními útvary státu Kentucky, ale taky s tzv. parsery, které už v titulech ze 70. let zpracovávaly hráčské příkazy v přirozeném jazyce. Dnes jsme v situaci, kdy všudypřítomní chatboti svedou mnohem víc, než jen vypsat správnou scénu na základě pokynu “jdi na západ”. Podíváme se na historické předchůdce a genealogii současného boomu velkých jazykových modelů nejen s ohledem na formální podobnost, ale taky logiku uživatelských rozhraní nebo funkci, kterou v budování fiktivních světů a tamních subjektivit hráče/uzivatele mají.
Diskuse s Karlem Pioreckým
18:30-19:00
Conference ending
19:00-19:30
Conference about Art, Literature, and Playfulness in the Context of Generative AI
29. - 30. 9. 2026
Praha, Kampus Hybernská - Didaktikon