Yesterday, Thursday, May 28, the Financial District campus of ITBA (Buenos Aires Institute of Technology) ceased to be a traditional academic space to become a living laboratory. Media Party University 2026 arrived with a premise as ambitious as it is necessary: the “Engineering of Truth.” The event achieved what few spaces manage to do: bringing to the same table those who write the news, those who train the algorithms, and those who study how our brains process both.
With a sold-out turnout, the day unfolded between the rigor of code, the analysis of bias, and the search for new models for a journalism that has already changed forever. The proposal was clear: to examine how technological systems intervene in civic reality and how, through an interdisciplinary approach, we can audit that mediation.
Keynotes: Perspectives from the Technological Vanguard
The morning session offered a roadmap on how major organizations are integrating artificial intelligence not just as a tool, but as a structural paradigm shift.
La Nación Data: The "Superpower" of Synergy and the Evolution of Data
Kicking off the first Keynote, Momi Peralta Ramos described the evolution of La Nación Data from its manual beginnings—where the team processed 100,000 election records and listened to 40,000 audio clips from the Nisman case—to the creation of the interdisciplinary “CROSS” teams and the AI Labs. This transition transformed data journalism into a “superpower” that now integrates technology, design, and knowledge to conduct complex investigations that were previously unattainable due to a lack of time and resources.
On the other hand, Martín Pascua detailed how AI currently redefines user experience and internal processes by cloning the voices of more than 20 journalists to humanize content, alongside animating the historical archive through the Flashbacks project. Furthermore, he highlighted the development of proprietary tools, such as conversational search engines for the newsroom, emphasizing that human oversight remains fundamental to mitigate errors or “hallucinations” from generative models.
Despegar: The Emergence of Large Consumer Models (LCM)
Pablo Abad, CTO of Despegar, presented Large Consumer Models (LCM), an evolution of AI that seeks to predict consumer actions instead of words. Trained on massive volumes of transactions, these models aim for proactive personalization that understands user needs after just a few clicks. According to Abad, although it is currently a “theoretical instrument,” current prototypes have already achieved a 50% improvement in bookings in specific campaigns.
The path toward mass adoption faces major investment barriers—exceeding 40 million dollars—and the requirement of massive datasets, a challenge that Despegar covers through the ProSus group. Abad concluded by warning about latency and the ethical risks of algorithmic bias, defining LCMs as the paradigm that will transform the industry in the coming years.
ITBA: Neuroscience and Cognitive Resistance
The presentation by Fabricio Ballarini (ITBA) was a warning from neuroscience regarding the vulnerability of our decision-making process when facing automation. The central point focused on how our brain, biologically designed to minimize energy expenditure, tends to delegate critical thinking to artificial intelligence, running the risk of accepting algorithmic biases or technological hallucinations as absolute truths.
Ballarini argued that the key is not the tool itself, but recovering the human capacity to ask the right questions. His core message was a call to cognitive resistance: faced with an AI that can fabricate bibliographies or perpetuate gender and race discrimination within data, humans must use technology to enhance real well-being and precise diagnostics.
Infobae: AI as Complementary Intelligence
Opy Morales, head of AI Strategy & Regional Operations at Infobae, detailed the creation of ScribNews, a platform for “complementary intelligence” that centralizes the use of agnostic models under its own style manual. Their strategy is based on “anti-prompting”, replacing the memorization of complex instructions with specific functions that automate processes such as SEO, metadata, and the translation of international wire services.
The implementation includes “Autonomouses,” a system of chained assistants that process long documents and perform fact-checking against original sources. Morales stressed that final responsibility always remains with the journalist, as the tool does not publish autonomously.
AI doesn’t transform companies. The employees who use it do,” he concluded.
Workshop Agenda: Taking Technical Action
In the afternoon, theory gave way to hands-on experimentation in the ITBA laboratories, organized around three strategic pillars:
- Generation Pillar: The activity began with a collective laboratory by David Coronel (LAIA) on tokens and vibe-coding to understand the computational cost of “thinking.” In parallel, Ernesto Mislej (7Puentes) provided tools to design AI initiatives, transforming strategic frameworks into executable skills.
- Absorption Pillar: Focused on how we decode technology. Juan Ignacio Galli (ITBA) explored the real limits of sentiment analysis through AI, while Ariel Tiferes (Chequeado) shared a session of “technical honesty” regarding the lessons and mistakes made during the creation of Sofía Fuentes, their synthetic avatar.
- Action Pillar: Addressed systemic impact. Roberto Abalde (ITBA) conducted a technical audit on biases within news recommendation systems, and Ariel Aizemberg (ITBA) closed with a creative look at the “grammar of juggling.”
A Necessary Thank You
The success of Media Party University 2026 is the result of a shared effort. We want to express our deepest gratitude to ITBA for opening its doors and establishing itself as the perfect ecosystem for this dialogue between engineering and journalism.
Thank you to the speakers for their transparency in sharing not only their achievements but also their ethical dilemmas and process failures, and to the participants—students, teachers, and professionals—who packed the venue and enriched every debate with their critical vision.
This day served as the ideal prologue for what is to come this October in our main edition. The Engineering of Truth is a process under construction, and yesterday we took a fundamental step toward understanding its foundations.
See you at the next Media Party!
