The poem “The Machine of the World” (A Máquina do Mundo, 1949) by the Brazilian poet Carlos Drummond de Andrade (1902-1987), written shortly after World War II, was chosen, in 2000, as the best Brazilian poem of all time, by a group of writers and literary critics at request of Caderno Mais of Folha de São Paulo newspaper. It is an enigmatic and dense text that requires a spiritual breath and at the same time a syntactic breath (Wisnik, 2022) in the course of its interpretation. The ancestral theme of the machine of the world, the one that the poet finds halfway through his journey, is formulated in this poem in a contemporary way: it is a machine that no longer offers itself to the modern and fragmented world as being capable of encompassing and giving visuality to the whole. The purpose of this presentation is to read the allegory of the machine of the world in Drummond's poem, as a compact or porous stone in the middle of the road, in dialogue with other poems and other times, from the 21st to the 13th century, and vice versa, from André Vallias (and his QR-coded diagram of Divine Comedy) to Dante Alighieri, from Dante to Camões, from Camões to Drummond, from Drummond to Haroldo de Campos, from Haroldo to Adriana Lisboa. In this hermeneutic path and not necessarily chronological, and based on the studies of Wolfgang Iser (The Act of Reading, the implied reader, and the meaning as a dynamic happening), I invite the audience to reflect on the issue of acts of literary reading in the contemporary world, the relevance of poetic and literary reading in the current context of the digital media and social networks, the poetry as the great machine of the world, passing through social, political, ecological, racial articulations, among others, inherent in the Portuguese-speaking world.

The growing realization that AI is not that intelligent after all has done little to dim the popular enthusiasm for and corporate promotion of its use. A desire for expediency fuels the former, whereas fantasies of workerless labor inform the latter. Both rely on assumptions about what constitutes sufficient quality output: the good enough. Whether the users of LLM-based text generators are able evaluators of the good enough depends on their standard of evaluation. Simple completion of a writing task could be enough. Evaluating the quality or efficacy of AI-generated prose does not lend itself to rapid, automatic assessment, defeating the purpose of employing AI in the first place. Although the mathematics of LLMs offer novel opportunities for machine translation and quantitative linguistics, their quotidian uses produce volumes of underread text for purposes that appear to be little beyond professional or academic obligations.

This paper investigates the ontological status of both the LLM as a creative agent and the generated text when employed to satisfy an uncritical standard of the good enough. Drawing on Anthropic’s work on interpretable AI, the paper argues that the strengths of LLMs are not aligned with the tasks to which users regularly put them. Though they can reveal, quantitatively, literary structures, they instead churn out fluent but often vapid prose suited to little purpose other than existing. Moreover, that existence is predicated on teleologies of writing that do not necessarily take communication as a goal—the good enough AI-generated text is accepted rather than read.

Recently digital technologies have entered into the fields of not only natural sciences but also humanities. Its literary evaluation apart, ChatGPT can write decent poems. With the help of a 3D printer, an AI program can create fine pieces of sculpture. Music that we casually listen to is now mostly done by DTM (desk top music). And we have song-writing AI programs, too. All this sounds extremely promising.

The purpose of this paper is to assess the achievements of the computer technologies in the spheres of poetry and music and delve into their socio-cultural significances. The reason that, out of many branches of humanistic activities, verse and music are specifically selected for discussion here is that these two genres of art are, I argue, closely connected both historically and formally.

The obvious Pro of the AI technologies in the humanistic creation is its ostensible high quality. The AI verses read beautifully. The AI songs sound pleasantly.

Here, however, immediately lies the pitfall of AI-generated pieces of art. They must depend on the existing conventions of art and the protocols of interpreting them shared by the receivers (that AI can detect and learn by scrutinizing the digital data on the net). Joyce first offended readers. Stravinsky at the premiere of Rites of Spring shocked and repelled audience. That is the fate of truly original art. AI technology can never scandalize the audience and, thus, create something truly original.

Secondly, the problem of AI’s incompatibility with semiotic ambiguity has to be pointed out. Essentially, AI technology is bound by the transparent signification as that is the feature of “normal” non-literary discourse which almost exclusively constitute the mega-data in the cyber space that AI relies on. AI cannot speak metaphorically, which is a significant setback of AI poetry.

This leads to the third problematic of AI technology: its logocentricity. Signification in AI-woven discourse cannot be but determinant. To use Bakhtinean terminology, the production of AI technology is always “finalized.”

These three problems seriously restrict the scope of AI-produced “literary” output. AI technology, however, in the reverse way, allows us to see the true essence of human cultural activity. In my paper I shall try to demonstrate the above stated points by analyzing, by way of an example, a poem by Maria Tsvetaeva and its song version in comparison with AI-produced verses and songs.

This paper proposes the study of the campus novel as a global genre, exploring the global reach and emergence of the genre in different literary traditions. The campus novel has historically been seen as a static and exclusive literary genre, resistant to change and prominent only in the British and North American literary tradition. Constant critical evaluation of only the Anglo-American tradition of the genre has then led to its impending demise, due to the lack of critical and mainstream attention towards a seemingly obsolete genre. As a result, scholars have continuously considered ways of revitalising the genre, initially by calling for diversity in campus novels, particularly following the success of Brandon Taylor’s Booker-shortlisted novel Real Life (2020). My research proposes the study of the campus novel as a global genre, exploring the different ways in which the genre can be found in literary traditions around the world. My study has found examples of the genre emerging in different cultures, through examples of the campus novel in cultures and locations as diverse as Germany, Norway, South Africa, and Indonesia, to name a few. There is also evidence of campus novel traditions that have emerged independent of the influence of the Anglo-American tradition, and existing under names of their own, such as the overseas student literature tradition in Taiwanese literature. These examples are in addition to campus novel traditions that have been acknowledged in studies of the genre but which continue to be deemed secondary to the Anglo-American tradition, as is the case of the campus novel in India and Egypt. In recent years, we have also seen resurgence of the genre through publication of acclaimed novels such as R.F. Kuang’s Babel (2022) and Elaine Hsieh Chou’s Disorientation (2022). The genre has also recently found new life through the contemporary popularity of various online phenomena, such as renewed interest in Donna Tartt’s The Secret History (1992) following the online rise of the aesthetic concept of dark academia. This paper thus promotes the campus novel genre as having the capacity to morph and evolve, evident in its emergence in various literary traditions and cultures around the world. This further challenges existing debates pronouncing the death of the genre, and considers the genre as contributing to existing studies of circulation of genres and literary globalisation. In addition, this paper also considers the state of higher education today, considering current trends and concerns, and how these may have led to contemporary interest in the literary genre.

The past two decades have seen a surge in so-called erasure, blackout, and other materially pronounced forms of poetry, with many writers cutting out, blackening, painting or stitching over existing texts as a way of engaging with them to tell other stories: stories that could not have been told or surged otherwise. These rewritings and overwritings can be seen as a form of cultural memory: they are practices that change the past as it is reassembled and shared in the present. Using different erasure technologies, the poems enable new experiences and forms of subjectivity, while highlighting the materiality of writing and re-writing. We propose to call such technologies ‘palimpsesting.’

In this paper, we discuss a number of recent texts that employ various erasure technologies, including Zong! and Nets as well as Insta poetries, exploring how meaning and materiality are entwined and what the agency of poetry is as reworked materiality.