Media Party New York

The Amazônia Vox project is supported by the ICFJ through the Knight Fellow program. It is developing a knowledge source bank of the Amazon, a network of communication freelancers from the Amazon, and content based on Solutions Journalism. It is also developing an image bank with productions by photographers and filmmakers from the region, including indigenous and traditional peoples. These are useful tools to help climate journalists or those interested in the region find reliable sources and professionals to support collaborative work, facilitating connections and generating opportunities and income in the territory.

I am proposing Bangla AI, an innovative AI-powered tool aimed at transforming the Bengali ethnic media landscape in the United States, specifically designed for Bengali journalists working in ethnic media in New York City. The project leverages generative artificial intelligence (GenAI) to support journalists in language translation and summarization, ensuring they can access and produce news in their native language with accuracy and speed. Who uses it and why? The primary users of Bangla AI are Bengali ethnic media journalists, especially those working in New York City, where a large Bangladeshi immigrant community resides. These journalists face significant language barriers when trying to gather, translate, and report mainstream news for their audience. Traditional tools like Google Translate often fall short in delivering contextually accurate translations, which compromises the quality of news. Bangla AI solves this issue by offering a more robust, contextually aware translation tool, allowing journalists to provide relevant and accurate news to the Bengali-speaking diaspora. Additionally, Bangla AI offers a summarization feature that condenses lengthy news articles into concise summaries in Bangla. This feature caters to the evolving reading habits of the audience, ensuring that journalists can keep their readers informed without overwhelming them with lengthy content. Value and Innovation: Bangla AI holds great value for me as it addresses a critical need in the Bengali ethnic media community, where language barriers often hinder journalists from providing accurate and relevant news. The innovative use of generative AI in this context marks a significant advancement in how ethnic media journalists can access, translate, and present information. By integrating AI with language translation and summarization, Bangla AI is bridging the gap between mainstream and ethnic media, empowering journalists to improve the quality and relevance of their news coverage. Impact: The impact of Bangla AI is far-reaching. It strengthens the role of ethnic media as a trusted source of news for the Bengali-speaking diaspora, allowing journalists to connect with their audience more effectively. It also addresses a critical language barrier, ensuring that vital information reaches non-English-speaking Bangladeshi immigrants in the U.S., particularly in New York City. As the project continues to evolve, it has the potential to extend its reach to other ethnic media communities facing similar challenges, enhancing the inclusivity and accessibility of media coverage for non-English-speaking populations. In summary, Bangla AI is an innovative, impactful tool that empowers ethnic media journalists, particularly in the Bengali-speaking community in New York City, to bridge language gaps, provide accurate news, and foster stronger connections with their audience.

Bibliofilia is an online platform dedicated to Spanish-language literature. It features in-depth interviews with prominent Latin American authors and English-speaking authors whose works intersect with Latin American themes. Our interviews explore authors' writing routines and tips on craft, as well as the significance of their work. Bibliofilia also provides book recommendations, curated insights on new releases, and interactive activities, including virtual and in-person book clubs, for Spanish speaking communities across the Americas. Recently, Bibliofilia won second place at the Media Party University in Buenos Aires for its innovative 'literary map' project, an interactive experience that uses programming to identify bookstores recommended by writers across the American continent, showcasing a creative integration of technology with media platforms and literary exploration. The platform has garnered a significant readership since it was launched 4 months ago, particularly among Spanish speakers in the United States, connecting readers with authors and fostering a vibrant literary community. Very few major literary publications include space for literature in Spanish, and as the number of Spanish speakers in the United States steadily increases, we hope to fill that gap. Looking ahead, Bibliofilia aims to evolve into a hub for Latin American literature, uniting readers and writers across the continent in a community of people connected by Hispanic heritage and love for literature. By offering original content on Latin American books, supporting authors, and promoting conversations, Bibliofilia aspires to bridge regional gaps and enrich cross-cultural understanding of Spanish-language literature, contributing to a deeper appreciation of contemporary literature and its cultural impact.

We worked together with El Confidencial in Spain and with VoxEurope in Belgium to personalize a Data Visualization recommendation solution that works with the most common data sources for non technical journalists who want to include charts in their reporting. We would like to showcase the design process and the results.

Investigative Toolkit is a beta web app featuring tools to help investigators to execute crucial but laborious technical tasks—such as translating and OCRing documents, using large language models, downloading videos, and image searching. The app is highly innovative and would build on open-source tools and also use the latest in Document Intelligence, Large Language Model, and other AI services. This would massively increase investigators’ speed, efficacy, and security. It would feature an elegant user interface, making it more accessible to investigators who are not technically savvy. The project could be a traditional SaaS setup or, more excitingly, could reside in the user's own cloud environment. If in the user's own cloud environment, it would take advantage of locked-down private endpoints to all services (from data storage to AI), meaning that data would not leave the Microsoft/Amazon/Google backbone until requested through the web app via peer-to-peer (the successor to VPN technology) access by investigators; this would eliminate infrastructure-based security risks. Girish has a unique background for this task. He was an investigative journalist for just under a decade, based primarily in Venezuela during the worst of its humanitarian crisis; he also worked across the Americas and Middle East, including in Iraq, Afghanistan, Egypt, and Cuba with everyone from the New Yorker to Reuters, where he was a Senior Correspondent. Girish is currently CTO of a Silicon Valley venture-backed startup, building enterprise software and architecture for the audit industry; in that role, he creates analogous SOC-2 compliant AI infrastructure in clients’ cloud environments. He also has a deep theoretical understanding of AI, having studied physics at Cambridge and Manchester, and machine learning at Stanford. He is therefore uniquely and highly qualified on both the technical and investigative dimensions of this project.

The MVA team presents the Media Viability Accelerator! An online platform to track, benchmark and optimize independent media outlets’ business performance, globally. The initiative is led by Internews, Microsoft and USAID, with a growing number of media support organizations supporting its development - as well as an expanding pool of media outlets putting their trust in the capabilities of the platform. MVA Insights (a business data tracking platform for media outlets) and MVA Solutions (an online marketplace of ideas and AI integrations) will be showcased by the MVA team, in demos of the platform and through vibrant conversations with booth visitors. Media Feria participants will learn more about opportunities to collaborate on this innovative media support initiative, how to accessSo  funding for technologists and media support organizations of up to 100,000 USD.

Privacy policies are long, but AI can read them in a second. Feeding in privacy compliance standards, LLM models like GPT can grade and visualize them, saving the pain of hard reading. PrivaSee is an AI-driven tool to decode obscure privacy policies and provide summarized reports, professional scores, and intuitive visualizations. It also archives privacy policy updates, visualizes changes between versions, and preserves a database for media, researchers, legal analysts and so on. More than just simplifying the complex, PrivaSee holds apps accountable for their data collection practices and safeguards your privacy in the age of transparency. This tool offers a platform that visualizes app policy statuses across different categories and countries, enabling readers to search on their own. For journalists covering technology, this is a tool that tracks updates in app privacy policies and alerts newsrooms for a breaking news story when updates occur.

In 2023, I founded SAOViVO, a free and open-source software that operates by extracting videos from YouTube, organizing them into a queue, and sending them as a live stream back to the platform. This project was picked by Google News Initiative's (GNI) Latin America challenge By leveraging SAOViVO, newsrooms can maximize their storytelling potential within constrained budgets, enabling them to allocate financial resources toward expanding coverage and reporting from their local communities, rather than investing heavily in studio infrastructure.

Sutita Rimay es una plataforma que lucha contra la desinformación. Nuestra machine learning procesa los componentes de una noticia y el evaluador identificará la veracidad o falsedad de la información. ¿Por qué usarlo? Método Nuestros algoritmos escanean el link de una información y nuestro evaluador identifica los criterios objetivos obteniendo así un porcentaje de veracidad o falsedad de la noticia en segundos. Rapidez Para los periodistas les urge una herramienta eficaz y rápida para difundir información. Para los publicistas, agencias, Dircom, broadcasting, es indispensable cuando ven perjudicada su reputación. Comunidad Para el usuario común es una herramienta que puede impedir la difusión de desinformación con acción del mismo ciudadano. Mediante la plataforma puede advertir a su sociedad sobre una noticia engañosa.

A deepfake detection platform/tool for political mis/disinformation. A non-profit with a top-class technical infrastructure and team. On the back-end, we aggregate state-of-the-art deepfake detection models created by academic institutions around the world and for-profits. Users then link to any social media link or upload their own media file into our tool and we provide a detailed analysis of how manipulated the media is. No other platform/tool takes our aggregation approach, and we hope to be a critical source for truth and media manipulation debunking for the upcoming 2024 election season - both in the US and internationally. Our goal is to support various media organizations at different scales, using AI for good.

The election of George Santos to Congress—despite his fraudulent background—raised serious questions about the media's inability to vet a new wave of political candidates with unknown backgrounds. This case inspired the creation of Sunlight Research Center, a non-partisan non-profit helping local newsrooms scrutinize 2024 candidates. In this keynote I will share the lessons Sunlight has learned about the real stories hiding behind the campaign biographies, the tech needed to enable deep investigation, and what hard-pressed newsrooms require in order to do this accountability reporting.

The 2023 elections in Nigeria experienced the first use of significantly new technology, directly connected to the election process itself but also coinciding with the proliferation of media-manipulating tools such as AI-generated deep fakes. Some examples include paid posts reportedly linking candidates to militant or separatist groups which filled social media platforms. The Nigeria presidential election was an instance of a country’s democratic process being impacted by such means. The scope of the problem was beyond the resources of local journalists to address. With wider access to more divergent technology, I believe the US can gain important risk and mitigation learnings from the critical role media tech played in Nigeria's 2023 elections. Learning: The session will explore how media innovation through the use of deeply insightful, investigative data stories and more effective fact-checking and deepfakes-exposing tools exposed these schemes and set a precedence for what might be improved voter understanding of media manipulation tricks and how to avoid it on an individual and community level.

What I have learned investigating AI impact on 2024 elections

Answering: How has the space evolved for both AI and hallucination detection. How GPTZero has grown and learned from both writers, journalists, and enterprise customers in adopting AI responsibly and keeping our information ecosystems safe.

In an industry set on rapid information flow and shortened attention spans, generative AI offers a powerful tool to enhance the value of journalism, particularly during the high-stakes environment of an election. By leveraging AI to delve into archives, we provide crucial historical context, draw meaningful connections between past and present events, and set a more relational tone for public discourse that people can make sense of. Moreover, the integration of generative AI with human-centric interactions, social listening, location data, sentiment analysis, and other indicators (economy, employment, health…) for each community, creates a multidimensional view of the election context is specific areas that makes it easy for the user to find familiar ground from where to absorb the information. This synthesis of information allows journalists to uncover patterns, anticipate trends, and provide more comprehensive coverage tailored to specific demographics or regions which also allows for richer, deeper and longer interactions. By harnessing these technological capabilities, we can take election journalism beyond the “we and them” horse-race reporting or some outlandish quote, offering citizens in-depth, contextualised information that they can relate and build upon from, so that they can make an informed decision at the ballot box.

Using the case on Zac Brettler's visual investigation as an example. This talk will delve into how data-driven visual storytelling, how we combined new evidence with intuitive visual storytelling to bring back to life Zac’s case and how a multidisciplinary team was key for the end result. By combining new evidence and data and fact checks with intuitive visual storytelling, this project aimed to take the user step by step on how that night unfolded allowing them to engage with the information on a personal level and where in the final result collaboration was key.

I want to highlight the inequality on Wikipedia. The Wiki community has been working hard to solve for this problem: one Wikiproject at a time.