Writing · Psychology

Introduction — Part 1 — A psychologist decides to become a writer using AIs — a universe as a latent narrative.

On the challenges and obstacles of writing good books — how AI and psychology came together.

Finally, when AI tools arrived to assist with writing, I faced the same flaws and errors as any novice writer — even with an AI that could help, though perhaps not quite enough yet, since early AIs weren't all that advanced. I struggled with stalling, a lack of creativity, writer's block, and other obstacles that plague writers of all kinds. Meanwhile, the AIs had their own shortcomings — ranging from hallucinations and originality issues to excessive superficiality or hidden artificiality — as well as difficulties interpreting requests, poor prompt responses, factual errors, and unfounded guesswork.

We write with AI because we believe it offers a gradual, educational learning experience in literary writing. We anticipate that when we eventually write books on our own, we will no longer rely on those "mentor" models — models we interact with under the illusion that they are crafting a story for us, when in reality, we end up doing three times the work (or more) just to fix their lack of originality and write like true authors, even while using AI as a starting point.

I have always aspired to be a great writer. I already have good books tucked away — works created over time without AI, long before it existed — and excellent material set aside for future publication. When AI arrived, I already had finished manuscripts that the technology helped me refine before I had even fully completed them. And this was before AI became truly proficient at the tasks it performs today — writing, editing, providing information, and grounding literature — capabilities that exist now, even if not every AI is quite ready for the job. I delved deeply into each of them; it was necessary to be honest and say, "this one works for writing, this one isn't ready." We concluded that GPT — from the beginning through phase 5 — remains the best for screenplays and for envisioning interwoven layers of fiction with rich narrative beats; however, it didn't quite cut it for crafting masterful paragraphs. While it wrote well, the output was dry, shapeless, and overly artificial. Google's tools were always confusing, though they did occasionally yield good paragraphs; nowadays, however, it is possible to write with one of them — Gemini, the latest release (writing this in 2026), is quite capable. The fact remains, though, that none of the tools tested — Qwen, Manus, Mistral, or DeepSeek — are truly ready to assist with literature yet. That said, one of them performed remarkably well: the Claude 4.6 models and beyond — both Opus and Sonnet — were indeed more capable than the rest. Of course, by tomorrow — as we turn the next corner or venture down new paths — they will all have surpassed their current limits. It won't just be Opus or Sonnet writing well enough for literature; they will all evolve to claim their place as solid tools for literary writing — and perhaps even generate entire books in one go.

Fortunately, we had the chance to examine each of them from the start. Experimenting with all of them gave us valuable insight and a learning foundation regarding how they were built. Currently, we rely primarily on GPT, Claude, and Mistral — the latter proving to be much better for literature over time. That's right — Mistral stepped up its game, even if it hasn't quite reached the level of Claude and its models yet.

Right now, we are testing Fable 5 — Anthropic's latest model — which is already showing signs of progress. …

I have always been connected to literature; from an early age, I would lose myself in books and stories — ranging from children's classics to the great classics of humanity by the time I reached adolescence. I was able to read the very best of the best. I always dreamed of being a writer — even though acting and theater served as my strongest artistic bridge, along with a touch of music in bands that eventually disbanded. Literature has always gone hand-in-hand with my personal development, and I have never abandoned it! As I will mention later, I also worked in bookstores, organizing and promoting events such as poetry readings, saraus (literary-musical gatherings), and storytelling sessions. I currently live in Valença, Rio de Janeiro, up in the highlands near the Serra das Araras range — which connects us to the city of Rio de Janeiro. Valença is widely known for being neighbors with Conservatória, a place where I've spent many weekends with friends. It is the state's most poetic refuge — a place where short stories and poetry blend with music and saraus. It is the city of saraus — the oldest in Brazil and the state of Rio de Janeiro — long known as a sanctuary and a place of rest for weary citizens from the city of Rio and the greater capital area.

However, I qualified as a psychologist — specifically an analytical psychologist — focusing on archetypes, complexes, and the collective unconscious.

My narrative focus is human-centric, revolving around myths, fairy tales, and clinical — as well as literary — studies. These support the notion that the world is a collective projection of stories and narratives, a place we inhabit alongside signs (language, linguistics), archetypes, and other surprising, unusual elements. My own name represents a convergence with a narrative universe far removed from my own — given that I am not Israeli, though I might have Jewish ancestry, considering the historical strength of Judaism in Portugal (such as in Lisbon or other locales). My name is Miguel Arcanjo Lima Maia. It stands as a testament to the fact that Brazil itself is a multifaceted, plural, and somewhat unique cultural tapestry — a narrative of names, mixed heritages, lessons learned, tales, and stories. A universe — inherently narrative or linguistic — that flows through all of us from a very early age. Being an archetype-focused psychologist — with a specialization and master's degree in analytical psychology, archetypes, and myths — has allowed me to see this more clearly. I have also taught extension courses on Nise da Silveira and lectured on these themes at universities, cultural centers, and cinemas. Since graduating, I have been an active volunteer lecturer, appearing at institutions such as Unitau, UFRRJ, UFRJ, FASF, USS-Vassouras, and UNISAL-Lorena — always presenting interesting talks or projects as a practicing psychologist. I have likely participated in other events and opportunities at bookstores and various venues not listed here. I have also lectured at events at UFMG (Belo Horizonte) and Uninove, participating in psychology gatherings, congresses, symposia, and other occasions to share or distill knowledge (though these last two instances occurred before I graduated).

But let's take things one step at a time. I am a psychologist with a postgraduate degree in Analytical Psychology from the Paeeon Clinic in São José dos Campos (affiliated with the UNISAL University Center in Lorena). I completed my lato sensu training under excellent Analytical Psychology professors from the state of São Paulo and the Paraíba Valley; one of them was Dr. Zacharias — author of Quati (a famous psychological test based on Jungian typology used worldwide) and other books. I consider my training in São José dos Campos — comprising analysis sessions, final course monographs, and other activities — to be among the most enriching experiences of my career. I later moved to the interior of Rio de Janeiro state — specifically the region known as the "Coffee Valley" (Vale do Café) — to live near a country estate featuring an old colonial-style house (with Dutch-style summer corrals still located below the main residence). There, I continued my work in psychology by teaching courses on personality theories, psychological assessment, testing, and psychometrics at the Faculdade Sul Fluminense (now FASF) in Volta Redonda. I also pursued further academic training in Seropédica at the Federal Rural University of Rio de Janeiro (UFRRJ) under the mentorship of Professor Dr. Nilton Sousa Santos, where I completed my master's degree (stricto sensu training) at this beautiful, expansive federal campus, which features lakes, trees, and gardens spread across the grounds surrounding the buildings. My work focused on a theme that had engaged me since my specialization: identifying narratives and potential narrative patterns within the local community and collective, more "tribal" behaviors — such as the psychology of drug use and chemical dependency, or social groups like the "Black Blocs." I still aim to produce books and articles that apply this same theoretical capacity to interweave psychological myths, rites, symbols, and archetypes with the public behavior of groups ranging from the Islamic State (ISIS) to drug trafficking factions — but it is not just about that, or about the "terrorism as narrative" characteristic of postmodern times, with the weakening of certain classic structures of identity. I worked as a psychologist at a CRAS center in Minas Gerais and as a coordinator for a CREAS unit — in the southern region of Rio de Janeiro state — always operating within the narrative paradigm and narrative psychology; in doing so, I observed residents, existing groups, and groups I had formed myself. I already had prior experience in bookstores — I actually worked in them — as well as in street theater and amateur theater; I had even acted and directed, hosted poetry and sarau gatherings, and organized literature and storytelling events for a local bookstore. I owe much of my theoretical background to reading icons such as James Hillman, Marie-Louise von Franz, C. G. Jung himself, and Luigi Zoja — Luigi Zoja, in particular, is able to articulate how collective unconscious, anthropological, and ancestral structures culminate in behaviors that are narrative in nature — in tribal and anthropological-archetypal terms.

Observing the population — their attire in public services, the struggle for diversity, and the class struggle for a more emancipated social-collective space — led me to notice narrative behaviors, conflicts, and dilemmas. These elements formed something akin to a narrative literature of humanity — its roles, its comings and goings, and life on the streets for both the poor and the middle class. I became convinced that a narrative quality permeated everything I did in my work as a psychologist dedicated to social services — a role that never caused me to lose sight of psychology itself. My interest in narrative anthropology intensified during this period. I even completed half of a history degree via distance learning at Unirio — another federal university — though I did not finish the program. However, I gained privileged experience through internships at historical documentation centers in Piraí, Rio de Janeiro — an old city renowned for its history and coffee plantations. There, I accessed documents that were simultaneously "narrative and historical" and listened to the stories of the workers at the center, where our main task was digitizing that history to create a digital archive for Piraí. As a professor of personality theories at colleges in my state, I was able to validate these observations with greater clarity and theoretical depth. Life, therefore, seemed like a living narrative — a constant movement back and forth; people came and went, yet there seemed to be a common thread connecting it all, along with a behavioral guide that was profoundly "narrative" in nature, serving as the hook for their actions. I accumulated years of clinical experience, practicing in various offices and clinics across the Rio-São Paulo corridor… I am, in fact, a native of the region known as the Paraíba Valley; I was born in Pindamonhangaba, São Paulo — located just a few kilometers from the city of São Paulo and, likewise, always just a few kilometers from the river.